[WI 2014]Context Recommendation Using Multi-label Classification
2 likes3,190 views
This document proposes a new type of recommender system called a context recommender that recommends appropriate contexts (e.g. time, location, companion) for users to consume items. It discusses how context recommenders are different than traditional and context-aware recommenders. It also presents the framework for context recommenders including algorithms using multi-label classification to directly predict contexts. The document reports on experiments comparing these algorithms on several datasets and finds that personalized algorithms outperform non-personalized ones and that certain multi-label classification algorithms like label powerset using support vector machines achieve the best performance.
1 of 23
Downloaded 43 times
![Intro – Type of Recommendations
• [Item Recommendations]
• [User Recommendations]](https://image.slidesharecdn.com/slidewic2014context-140805210546-phpapp01/85/WI-2014-Context-Recommendation-Using-Multi-label-Classification-3-320.jpg)
![References
• Context Recommendations
[1].Baltrunas, Linas, Marius Kaminskas, Francesco Ricci, Lior
Rokach, Bracha Shapira, and Karl-Heinz Luke. "Best usage
context prediction for music tracks." In Proceedings of the 2nd
Workshop on Context Aware Recommender Systems. 2010.
[2].Yong Zheng, Bamshad Mobasher, Robin Burke. "Context
Recommendation Using Multi-label Classification". In Proceedings
of the 13th IEEE/WIC/ACM International Conference on Web
Intelligence, 2014
• Multi-label Classifications
[1].Tsoumakas, Grigorios, and Ioannis Katakis. "Multi-label
classification: An overview." International Journal of Data
Warehousing and Mining (IJDWM) 3, no. 3 (2007): 1-13.
[2].Tsoumakas, Grigorios, Ioannis Katakis, and Ioannis Vlahavas.
"Mining multi-label data." In Data mining and knowledge
discovery handbook, pp. 667-685. Springer US, 2010.](https://image.slidesharecdn.com/slidewic2014context-140805210546-phpapp01/85/WI-2014-Context-Recommendation-Using-Multi-label-Classification-21-320.jpg)
![[WI 2014]Context Recommendation Using Multi-label Classification](https://image.slidesharecdn.com/slidewic2014context-140805210546-phpapp01/85/WI-2014-Context-Recommendation-Using-Multi-label-Classification-23-320.jpg)
Ad
Recommended
[SOCRS2013]Differential Context Modeling in Collaborative Filtering
[SOCRS2013]Differential Context Modeling in Collaborative FilteringYONG ZHENG This document discusses differential context modeling (DCM) in collaborative filtering recommender systems. DCM is a framework that separates recommender algorithms into components and applies differential context constraints to each component to maximize contextual effects. The document applies DCM using differential context relaxation and weighting to item-based collaborative filtering and Slope One recommender algorithms. Experimental results on movie and food rating datasets show that differential context weighting improves predictive performance over baselines and differential context relaxation. Future work involves expanding DCM to additional recommender algorithms and optimizing performance.
[CIKM 2014] Deviation-Based Contextual SLIM Recommenders
[CIKM 2014] Deviation-Based Contextual SLIM RecommendersYONG ZHENG The document discusses the development and evaluation of deviation-based contextual slim recommenders (CSLIM) within context-aware recommender systems (CARS). It highlights the need for such models to consider various contextual factors and introduces methodologies for incorporating these contexts into collaborative filtering techniques. Experimental results demonstrate that CSLIM can outperform traditional recommendation algorithms, with future work suggesting improvements and adaptations for broader applications.
[ECWEB2012]Differential Context Relaxation for Context-Aware Travel Recommend...
[ECWEB2012]Differential Context Relaxation for Context-Aware Travel Recommend...YONG ZHENG The document discusses differential context relaxation for context-aware travel recommendation systems, emphasizing how contextual features can enhance personalized recommendations. It outlines various approaches, including filtering and modeling, and introduces the concept of differential context relaxation to optimize recommendation algorithms. By leveraging context-linked features and managing the tradeoff between accuracy and coverage, the authors explore methods for improving the relevance of travel recommendations based on user contexts.
[UMAP2013] Recommendation with Differential Context Weighting
[UMAP2013] Recommendation with Differential Context WeightingYONG ZHENG The document discusses a new approach to recommendation systems called differential context relaxation (DCR) and differential context weighting (DCW) that improves the handling of context sparsity in context-aware recommender systems. Through experimentation, the authors demonstrate that DCW significantly outperforms DCR and traditional methods in predictive accuracy while managing context sparsity effectively. Future work includes exploring alternative similarity measures and optimizing the algorithm further using parallel processing.
[CARS2012@RecSys]Optimal Feature Selection for Context-Aware Recommendation u...
[CARS2012@RecSys]Optimal Feature Selection for Context-Aware Recommendation u...YONG ZHENG This document summarizes a research paper on optimal feature selection for context-aware recommendation systems using differential relaxation. The paper proposes a differential context relaxation (DCR) model that applies different context relaxations to different components of a recommendation algorithm to maximize their contributions. It uses binary particle swarm optimization to efficiently find optimal context relaxations and outperforms exhaustive search. Experimental results on a food preference dataset show the effects of different contexts and context-linked features. The paper discusses limitations and opportunities for future work to address sparsity issues.
[SAC2014]Splitting Approaches for Context-Aware Recommendation: An Empirical ...
[SAC2014]Splitting Approaches for Context-Aware Recommendation: An Empirical ...YONG ZHENG This document describes an empirical study that compares different context-aware recommendation approaches. It evaluates three context-aware splitting approaches (item splitting, user splitting, and UI splitting) on several datasets using different recommendation algorithms and impurity criteria for splitting. The results show that UI splitting generally performs the best when used with matrix factorization as the recommendation algorithm. The document also compares the splitting approaches to other context-aware recommendation methods like differential context modeling and context-aware matrix factorization. The goal is to better understand how different context-aware techniques compare and which may be most appropriate depending on the data and application.
[SAC 2015] Improve General Contextual SLIM Recommendation Algorithms By Facto...
[SAC 2015] Improve General Contextual SLIM Recommendation Algorithms By Facto...YONG ZHENG This document summarizes a research paper that improves on a previous context-aware recommender system algorithm called GCSLIM by factorizing contexts to address its sparsity problem. The paper introduces GCSLIM and its drawback of measuring context deviations in pairs, which can result in unknown deviations when new context combinations are encountered. To solve this, the paper represents each context as a vector and calculates deviations as the Euclidean distance between vectors. Experimental results on a restaurant dataset show improved precision and MAP over baselines. The conclusions discuss how factorizing contexts can alleviate but not fully solve sparsity, and future work to address cold start issues.
[WISE 2015] Similarity-Based Context-aware Recommendation
[WISE 2015] Similarity-Based Context-aware RecommendationYONG ZHENG This paper proposes a similarity-based approach for contextual modeling in context-aware recommender systems. It introduces three methods for representing context similarity - independent, latent, and multidimensional - and applies them to context-aware matrix factorization and sparse linear models. Experimental results on four datasets show the multidimensional context similarity approach outperforms deviation-based contextual modeling and independent context modeling. The paper concludes similarity-based contextual modeling provides a general way to incorporate contexts and recommends exploring solutions to reduce costs in multidimensional modeling and applying other base recommender algorithms.
Download
Downloadbutest The document discusses several collaborative filtering techniques for making recommendations, including k-nearest neighbors (kNN), naive Bayes classification, singular value decomposition (SVD), and probabilistic models. It provides examples of how these methods work, such as using ratings from similar users to predict a user's rating for an item (kNN), and decomposing a ratings matrix to capture relationships between users and items (SVD). The techniques vary in their assumptions, complexity, and ability to incorporate additional user/item metadata. Evaluation on new data is important to ensure the methods generalize well beyond the training data.
Recommender Systems: Advances in Collaborative Filtering
Recommender Systems: Advances in Collaborative FilteringChangsung Moon This document summarizes recommender systems, focusing on collaborative filtering techniques. It discusses how recommender systems help with information overload by matching users with relevant items. Collaborative filtering is introduced as a technique that seeks to predict user preferences based on other similar users' ratings. The document then covers various collaborative filtering algorithms like neighborhood models, latent factor models using matrix factorization, and extensions like adding biases and temporal dynamics. It concludes by discussing hybrid methods and providing references for further reading.
Matrix Factorization Techniques For Recommender Systems
Matrix Factorization Techniques For Recommender SystemsLei Guo The document discusses matrix factorization techniques for recommender systems. It begins by describing common recommender system strategies like content-based and collaborative filtering approaches. It then introduces matrix factorization methods, which characterize both users and items by vectors of latent factors inferred from rating patterns. The basic matrix factorization model approximates user ratings as the inner product of user and item vectors in the joint latent factor space. Learning algorithms like stochastic gradient descent and alternating least squares are used to compute the user and item vectors by minimizing a regularized error function on known ratings.
Machine learning Algorithms with a Sagemaker demo
Machine learning Algorithms with a Sagemaker demoHridyesh Bisht An algorithm is a set of steps to solve a problem. Supervised learning uses labeled training data to teach models patterns which they can then use to predict labels for new unlabeled data. Unsupervised learning uses clustering and pattern detection to analyze and group unlabeled data. SageMaker is a fully managed service that allows users to build, train and deploy machine learning models and includes components for managing notebooks, labeling data, and deploying models through endpoints.
Matrix Factorization Technique for Recommender Systems
Matrix Factorization Technique for Recommender SystemsAladejubelo Oluwashina This document discusses matrix factorization techniques for recommender systems. It begins with an introduction to recommender systems and the content-based and collaborative filtering approaches. It then describes the matrix factorization model, which characterizes users and items with vectors to predict ratings. Methods like stochastic gradient descent and alternating least squares are used to optimize the model. The Netflix Prize competition is discussed, which helped popularize these techniques. In conclusion, matrix factorization has become dominant in collaborative filtering due to its superior accuracy over other methods.
Algorithms Design Patterns
Algorithms Design PatternsAshwin Shiv The document discusses various algorithms design approaches and patterns including divide and conquer, greedy algorithms, dynamic programming, backtracking, and branch and bound. It provides examples of each along with pseudocode. Specific algorithms discussed include binary search, merge sort, knapsack problem, shortest path problems, and the traveling salesman problem. The document is authored by Ashwin Shiv, a second year computer science student at NIT Delhi.
Evolving Reinforcement Learning Algorithms, JD. Co-Reyes et al, 2021
Evolving Reinforcement Learning Algorithms, JD. Co-Reyes et al, 2021Chris Ohk The document discusses the development of a new meta-learning framework for designing reinforcement learning algorithms automatically, aiming to reduce manual efforts while enabling the creation of domain-agnostic, efficient algorithms. The authors propose a search language based on genetic programming to express symbolic loss functions and utilize regularized evolution for optimizing these algorithms across various environments. They demonstrate that this approach successfully outperforms existing algorithms by learning two new algorithms that generalize well to unseen environments.
(Gaurav sawant & dhaval sawlani)bia 678 final project report
(Gaurav sawant & dhaval sawlani)bia 678 final project reportGaurav Sawant This report analyzes a movie recommendation system built using the MovieLens dataset, focusing on collaborative filtering techniques. Several methods, including Alternating Least Squares (ALS) and Pearson’s correlation coefficient, were evaluated for their performance in making accurate recommendations. Results indicate that while ALS shows decreased RMSE with increased data size, both Pearson’s and cosine similarity methods yield similar results, with cluster computing proving more efficient than local machines.
Recommendation and Information Retrieval: Two Sides of the Same Coin?
Recommendation and Information Retrieval: Two Sides of the Same Coin?Arjen de Vries The document discusses the relationship between recommendation systems and information retrieval, focusing on collaborative filtering (CF) and language modeling. It covers various approaches, including user-based and item-based CF, probabilistic modeling for user-item relevance, and challenges like data sparsity. The paper also explores the integration of CF and information retrieval techniques to improve recommendations and relevance ranking.
AI: AI & Problem Solving
AI: AI & Problem SolvingDataminingTools Inc The document discusses various problem solving techniques in artificial intelligence, including different types of problems, components of well-defined problems, measuring problem solving performance, and different search strategies. It describes single-state and multiple-state problems, and defines the key components of a problem including the data type, operators, goal test, and path cost. It also explains different search strategies such as breadth-first search, uniform cost search, depth-first search, depth-limited search, iterative deepening search, and bidirectional search.
Latent factor models for Collaborative Filtering
Latent factor models for Collaborative Filteringsscdotopen The document discusses latent factor models for collaborative filtering. It describes how latent factor models (1) map both users and items to a latent factor space to characterize them, (2) approximate ratings as the dot product of user and item vectors, and (3) can be used to predict unknown ratings. It also covers techniques like stochastic gradient descent and alternating least squares for training latent factor models on explicit and implicit feedback data.
Item Based Collaborative Filtering Recommendation Algorithms
Item Based Collaborative Filtering Recommendation Algorithmsnextlib The document summarizes research on item-based collaborative filtering recommendation algorithms. It analyzes techniques for computing item-item similarities and generating recommendations from the similarities. Experimental results show that item-based collaborative filtering provides better quality recommendations than user-based approaches, especially for sparse datasets. The regression-based prediction computation technique outperforms the weighted sum approach.
Collaborative Filtering at Spotify
Collaborative Filtering at SpotifyErik Bernhardsson The document discusses Spotify's use of collaborative filtering for music recommendations, focusing on techniques like matrix completion and latent factor models. It highlights the challenges of parallelization and explores various methods for measuring item similarity, such as cosine similarity and Pearson correlation. Additionally, the text touches on the need for scalable solutions in different domains, presents the importance of A/B testing, and concludes with a hiring note.
Deep reinforcement learning from scratch
Deep reinforcement learning from scratchJie-Han Chen 1. The document provides an overview of deep reinforcement learning and the Deep Q-Network algorithm. It defines the key concepts of Markov Decision Processes including states, actions, rewards, and policies.
2. The Deep Q-Network uses a deep neural network as a function approximator to estimate the optimal action-value function. It employs experience replay and a separate target network to stabilize learning.
3. Experiments applying DQN to the Atari 2600 game Space Invaders are discussed, comparing different loss functions and optimizers. The standard DQN configuration with MSE loss and RMSProp performed best.
Cs583 recommender-systems
Cs583 recommender-systemsAravindharamanan S The document summarizes recommender systems and different recommendation approaches. It introduces content-based recommendation which recommends items similar to those a user preferred. It also covers collaborative filtering which recommends items liked by similar users. Specific collaborative filtering techniques discussed include k-nearest neighbors, association rules, and matrix factorization using singular value decomposition as was done for the Netflix Prize contest.
A Multiscale Visualization of Attention in the Transformer Model
A Multiscale Visualization of Attention in the Transformer Modeltaeseon ryu The document discusses a multiscale visualization of attention in transformer models, including Transformer, BERT, and GPT, along with their respective architectures and tasks. It emphasizes the challenges of visualizing multi-layer, multi-head attention mechanisms and presents specific use cases for the visualization tool, such as detecting model bias and aiding in paraphrase detection. The conclusion suggests that while the visualizations are appealing, more detailed explanations of the weight extraction process would enhance understanding.
Using Interactive Genetic Algorithm for Requirements Prioritization
Using Interactive Genetic Algorithm for Requirements Prioritization Francis Palma This document describes using a genetic algorithm to prioritize requirements. It begins with an outline and introduction to the problem of prioritizing software requirements. It then discusses related prioritization techniques from literature and how they have limitations like poor scalability. The document proposes using a genetic algorithm to prioritize requirements, leveraging domain knowledge graphs representing priority and dependencies. It describes representing potential solutions as individuals in a population, calculating fitness by counting disagreements with domain knowledge, and using genetic operators like crossover to evolve better solutions over generations. The goal is to find a prioritized requirements list that best satisfies constraints and delivers value to users.
Evaluation of subjective answers using glsa enhanced with contextual synonymy
Evaluation of subjective answers using glsa enhanced with contextual synonymyijnlc This document presents a study on enhancing the evaluation of subjective exam answers using a software model based on Generalized Latent Semantic Analysis (GLSA) combined with contextual synonymy. It outlines the challenge of subjectivity in automated grading, the key techniques utilized in the model, and the improvements achieved over previous approaches, particularly in capturing varying expressions of the same entity. The system is implemented and tested using student responses, demonstrating successful automation and consistency in grading subjective technical answers.
Abstractive Text Summarization
Abstractive Text SummarizationTho Phan The document discusses the concepts of abstractive text summarization using sequence-to-sequence models with attention mechanisms, highlighting their role in natural language generation. It explains the differences between extractive and abstractive summarization, the importance of attention in improving model performance, and various decoding algorithms. Additionally, it covers evaluation metrics like ROUGE and challenges in measuring the quality of generated summaries.
Abigail See - 2017 - Get To The Point: Summarization with Pointer-Generator N...
Abigail See - 2017 - Get To The Point: Summarization with Pointer-Generator N...Association for Computational Linguistics The document describes a pointer-generator network for abstractive text summarization. It improves on sequence-to-sequence models by incorporating a pointing mechanism to copy words from the source text, addressing issues of reproducing factual errors or repetition. Coverage is also used to reduce repetition by penalizing attention to already attended parts of the text. Evaluation on the CNN/Daily Mail dataset shows the model outperforms previous abstractive methods according to ROUGE scores, achieving a ROUGE-1 of 39.5. However, summarization is difficult to evaluate and ROUGE has limitations as a metric. Future work includes improving long-form text understanding and generating more human-like summaries.
Recommender system algorithm and architecture
Recommender system algorithm and architectureLiang Xiang 1) The document discusses recommender system algorithms and architecture. It covers common recommendation techniques like collaborative filtering, content-based filtering, and graph-based recommendations.
2) It also discusses challenges like cold starts for new users and items. For new users, it recommends using demographic data or initial feedback to understand interests. For new items, it suggests using content information or initial user feedback.
3) The document proposes a feature-based recommendation framework that connects users, items, and latent features to address challenges like heterogeneous data and cold starts. This framework provides explanations but does not support user-based methods.
Tutorial: Context-awareness In Information Retrieval and Recommender Systems
Tutorial: Context-awareness In Information Retrieval and Recommender SystemsYONG ZHENG The document provides an overview of a tutorial on context-awareness in information retrieval and recommender systems. It discusses topics such as information overload, solutions like information retrieval (e.g. search engines) and recommender systems (e.g. movie recommendations). It then covers context and context-awareness, giving examples like how recommendations may change based on location, time, user intent, etc. It also discusses incorporating context-awareness into information retrieval and recommender systems to improve recommendations.
More Related Content
What's hot (20)
Download
Downloadbutest The document discusses several collaborative filtering techniques for making recommendations, including k-nearest neighbors (kNN), naive Bayes classification, singular value decomposition (SVD), and probabilistic models. It provides examples of how these methods work, such as using ratings from similar users to predict a user's rating for an item (kNN), and decomposing a ratings matrix to capture relationships between users and items (SVD). The techniques vary in their assumptions, complexity, and ability to incorporate additional user/item metadata. Evaluation on new data is important to ensure the methods generalize well beyond the training data.
Recommender Systems: Advances in Collaborative Filtering
Recommender Systems: Advances in Collaborative FilteringChangsung Moon This document summarizes recommender systems, focusing on collaborative filtering techniques. It discusses how recommender systems help with information overload by matching users with relevant items. Collaborative filtering is introduced as a technique that seeks to predict user preferences based on other similar users' ratings. The document then covers various collaborative filtering algorithms like neighborhood models, latent factor models using matrix factorization, and extensions like adding biases and temporal dynamics. It concludes by discussing hybrid methods and providing references for further reading.
Matrix Factorization Techniques For Recommender Systems
Matrix Factorization Techniques For Recommender SystemsLei Guo The document discusses matrix factorization techniques for recommender systems. It begins by describing common recommender system strategies like content-based and collaborative filtering approaches. It then introduces matrix factorization methods, which characterize both users and items by vectors of latent factors inferred from rating patterns. The basic matrix factorization model approximates user ratings as the inner product of user and item vectors in the joint latent factor space. Learning algorithms like stochastic gradient descent and alternating least squares are used to compute the user and item vectors by minimizing a regularized error function on known ratings.
Machine learning Algorithms with a Sagemaker demo
Machine learning Algorithms with a Sagemaker demoHridyesh Bisht An algorithm is a set of steps to solve a problem. Supervised learning uses labeled training data to teach models patterns which they can then use to predict labels for new unlabeled data. Unsupervised learning uses clustering and pattern detection to analyze and group unlabeled data. SageMaker is a fully managed service that allows users to build, train and deploy machine learning models and includes components for managing notebooks, labeling data, and deploying models through endpoints.
Matrix Factorization Technique for Recommender Systems
Matrix Factorization Technique for Recommender SystemsAladejubelo Oluwashina This document discusses matrix factorization techniques for recommender systems. It begins with an introduction to recommender systems and the content-based and collaborative filtering approaches. It then describes the matrix factorization model, which characterizes users and items with vectors to predict ratings. Methods like stochastic gradient descent and alternating least squares are used to optimize the model. The Netflix Prize competition is discussed, which helped popularize these techniques. In conclusion, matrix factorization has become dominant in collaborative filtering due to its superior accuracy over other methods.
Algorithms Design Patterns
Algorithms Design PatternsAshwin Shiv The document discusses various algorithms design approaches and patterns including divide and conquer, greedy algorithms, dynamic programming, backtracking, and branch and bound. It provides examples of each along with pseudocode. Specific algorithms discussed include binary search, merge sort, knapsack problem, shortest path problems, and the traveling salesman problem. The document is authored by Ashwin Shiv, a second year computer science student at NIT Delhi.
Evolving Reinforcement Learning Algorithms, JD. Co-Reyes et al, 2021
Evolving Reinforcement Learning Algorithms, JD. Co-Reyes et al, 2021Chris Ohk The document discusses the development of a new meta-learning framework for designing reinforcement learning algorithms automatically, aiming to reduce manual efforts while enabling the creation of domain-agnostic, efficient algorithms. The authors propose a search language based on genetic programming to express symbolic loss functions and utilize regularized evolution for optimizing these algorithms across various environments. They demonstrate that this approach successfully outperforms existing algorithms by learning two new algorithms that generalize well to unseen environments.
(Gaurav sawant & dhaval sawlani)bia 678 final project report
(Gaurav sawant & dhaval sawlani)bia 678 final project reportGaurav Sawant This report analyzes a movie recommendation system built using the MovieLens dataset, focusing on collaborative filtering techniques. Several methods, including Alternating Least Squares (ALS) and Pearson’s correlation coefficient, were evaluated for their performance in making accurate recommendations. Results indicate that while ALS shows decreased RMSE with increased data size, both Pearson’s and cosine similarity methods yield similar results, with cluster computing proving more efficient than local machines.
Recommendation and Information Retrieval: Two Sides of the Same Coin?
Recommendation and Information Retrieval: Two Sides of the Same Coin?Arjen de Vries The document discusses the relationship between recommendation systems and information retrieval, focusing on collaborative filtering (CF) and language modeling. It covers various approaches, including user-based and item-based CF, probabilistic modeling for user-item relevance, and challenges like data sparsity. The paper also explores the integration of CF and information retrieval techniques to improve recommendations and relevance ranking.
AI: AI & Problem Solving
AI: AI & Problem SolvingDataminingTools Inc The document discusses various problem solving techniques in artificial intelligence, including different types of problems, components of well-defined problems, measuring problem solving performance, and different search strategies. It describes single-state and multiple-state problems, and defines the key components of a problem including the data type, operators, goal test, and path cost. It also explains different search strategies such as breadth-first search, uniform cost search, depth-first search, depth-limited search, iterative deepening search, and bidirectional search.
Latent factor models for Collaborative Filtering
Latent factor models for Collaborative Filteringsscdotopen The document discusses latent factor models for collaborative filtering. It describes how latent factor models (1) map both users and items to a latent factor space to characterize them, (2) approximate ratings as the dot product of user and item vectors, and (3) can be used to predict unknown ratings. It also covers techniques like stochastic gradient descent and alternating least squares for training latent factor models on explicit and implicit feedback data.
Item Based Collaborative Filtering Recommendation Algorithms
Item Based Collaborative Filtering Recommendation Algorithmsnextlib The document summarizes research on item-based collaborative filtering recommendation algorithms. It analyzes techniques for computing item-item similarities and generating recommendations from the similarities. Experimental results show that item-based collaborative filtering provides better quality recommendations than user-based approaches, especially for sparse datasets. The regression-based prediction computation technique outperforms the weighted sum approach.
Collaborative Filtering at Spotify
Collaborative Filtering at SpotifyErik Bernhardsson The document discusses Spotify's use of collaborative filtering for music recommendations, focusing on techniques like matrix completion and latent factor models. It highlights the challenges of parallelization and explores various methods for measuring item similarity, such as cosine similarity and Pearson correlation. Additionally, the text touches on the need for scalable solutions in different domains, presents the importance of A/B testing, and concludes with a hiring note.
Deep reinforcement learning from scratch
Deep reinforcement learning from scratchJie-Han Chen 1. The document provides an overview of deep reinforcement learning and the Deep Q-Network algorithm. It defines the key concepts of Markov Decision Processes including states, actions, rewards, and policies.
2. The Deep Q-Network uses a deep neural network as a function approximator to estimate the optimal action-value function. It employs experience replay and a separate target network to stabilize learning.
3. Experiments applying DQN to the Atari 2600 game Space Invaders are discussed, comparing different loss functions and optimizers. The standard DQN configuration with MSE loss and RMSProp performed best.
Cs583 recommender-systems
Cs583 recommender-systemsAravindharamanan S The document summarizes recommender systems and different recommendation approaches. It introduces content-based recommendation which recommends items similar to those a user preferred. It also covers collaborative filtering which recommends items liked by similar users. Specific collaborative filtering techniques discussed include k-nearest neighbors, association rules, and matrix factorization using singular value decomposition as was done for the Netflix Prize contest.
A Multiscale Visualization of Attention in the Transformer Model
A Multiscale Visualization of Attention in the Transformer Modeltaeseon ryu The document discusses a multiscale visualization of attention in transformer models, including Transformer, BERT, and GPT, along with their respective architectures and tasks. It emphasizes the challenges of visualizing multi-layer, multi-head attention mechanisms and presents specific use cases for the visualization tool, such as detecting model bias and aiding in paraphrase detection. The conclusion suggests that while the visualizations are appealing, more detailed explanations of the weight extraction process would enhance understanding.
Using Interactive Genetic Algorithm for Requirements Prioritization
Using Interactive Genetic Algorithm for Requirements Prioritization Francis Palma This document describes using a genetic algorithm to prioritize requirements. It begins with an outline and introduction to the problem of prioritizing software requirements. It then discusses related prioritization techniques from literature and how they have limitations like poor scalability. The document proposes using a genetic algorithm to prioritize requirements, leveraging domain knowledge graphs representing priority and dependencies. It describes representing potential solutions as individuals in a population, calculating fitness by counting disagreements with domain knowledge, and using genetic operators like crossover to evolve better solutions over generations. The goal is to find a prioritized requirements list that best satisfies constraints and delivers value to users.
Evaluation of subjective answers using glsa enhanced with contextual synonymy
Evaluation of subjective answers using glsa enhanced with contextual synonymyijnlc This document presents a study on enhancing the evaluation of subjective exam answers using a software model based on Generalized Latent Semantic Analysis (GLSA) combined with contextual synonymy. It outlines the challenge of subjectivity in automated grading, the key techniques utilized in the model, and the improvements achieved over previous approaches, particularly in capturing varying expressions of the same entity. The system is implemented and tested using student responses, demonstrating successful automation and consistency in grading subjective technical answers.
Abstractive Text Summarization
Abstractive Text SummarizationTho Phan The document discusses the concepts of abstractive text summarization using sequence-to-sequence models with attention mechanisms, highlighting their role in natural language generation. It explains the differences between extractive and abstractive summarization, the importance of attention in improving model performance, and various decoding algorithms. Additionally, it covers evaluation metrics like ROUGE and challenges in measuring the quality of generated summaries.
Abigail See - 2017 - Get To The Point: Summarization with Pointer-Generator N...
Abigail See - 2017 - Get To The Point: Summarization with Pointer-Generator N...Association for Computational Linguistics The document describes a pointer-generator network for abstractive text summarization. It improves on sequence-to-sequence models by incorporating a pointing mechanism to copy words from the source text, addressing issues of reproducing factual errors or repetition. Coverage is also used to reduce repetition by penalizing attention to already attended parts of the text. Evaluation on the CNN/Daily Mail dataset shows the model outperforms previous abstractive methods according to ROUGE scores, achieving a ROUGE-1 of 39.5. However, summarization is difficult to evaluate and ROUGE has limitations as a metric. Future work includes improving long-form text understanding and generating more human-like summaries.
Abigail See - 2017 - Get To The Point: Summarization with Pointer-Generator N...
Abigail See - 2017 - Get To The Point: Summarization with Pointer-Generator N...Association for Computational Linguistics
Viewers also liked (20)
Recommender system algorithm and architecture
Recommender system algorithm and architectureLiang Xiang 1) The document discusses recommender system algorithms and architecture. It covers common recommendation techniques like collaborative filtering, content-based filtering, and graph-based recommendations.
2) It also discusses challenges like cold starts for new users and items. For new users, it recommends using demographic data or initial feedback to understand interests. For new items, it suggests using content information or initial user feedback.
3) The document proposes a feature-based recommendation framework that connects users, items, and latent features to address challenges like heterogeneous data and cold starts. This framework provides explanations but does not support user-based methods.
Tutorial: Context-awareness In Information Retrieval and Recommender Systems
Tutorial: Context-awareness In Information Retrieval and Recommender SystemsYONG ZHENG The document provides an overview of a tutorial on context-awareness in information retrieval and recommender systems. It discusses topics such as information overload, solutions like information retrieval (e.g. search engines) and recommender systems (e.g. movie recommendations). It then covers context and context-awareness, giving examples like how recommendations may change based on location, time, user intent, etc. It also discusses incorporating context-awareness into information retrieval and recommender systems to improve recommendations.
genetic algorithm based music recommender system
genetic algorithm based music recommender systemneha pevekar The document presents a genetic algorithm-based music recommender system (GAMR) that aims to generate meaningful recommendations and adapt to users' changing preferences. It combines conventional techniques with a genetic algorithm to improve recommendation accuracy and user interface engagement. The proposed system addresses the limitations of existing recommendation systems that often fail to reflect current user interests.
Matrix Factorization In Recommender Systems
Matrix Factorization In Recommender SystemsYONG ZHENG The document discusses matrix factorization techniques for recommender systems. It begins with an overview of recommender systems and their use of matrix factorization for dimensionality reduction. Principal component analysis and singular value decomposition are described as early linear algebra techniques used for this purpose. The document then focuses on how these techniques evolved into basic and extended matrix factorization methods in recommender systems, using the Netflix Prize competition as an example.
Recommender system introduction
Recommender system introductionLiang Xiang This document discusses recommender systems, including:
1. It provides an overview of recommender systems, their history, and common problems like top-N recommendation and rating prediction.
2. It then discusses what makes a good recommender system, including experiment methods like offline, user surveys, and online experiments, as well as evaluation metrics like prediction accuracy, diversity, novelty, and user satisfaction.
3. Key metrics that are important to evaluate recommender systems are discussed, such as user satisfaction, prediction accuracy, coverage, diversity, novelty, serendipity, trust, robustness, and response time. The document emphasizes selecting metrics based on business goals.
Building a Recommendation Engine - An example of a product recommendation engine
Building a Recommendation Engine - An example of a product recommendation engineNYC Predictive Analytics This document provides an example of building a predictive model for product recommendations. It outlines using a k-Nearest Neighbor (kNN) algorithm and singular value decomposition (SVD) for dimensionality reduction on order history and cart event data to create a recommendation engine. It discusses selecting features from the data, normalizing the data, using kNN to find similar items, and reducing the data dimensions with SVD before applying kNN. It also introduces using a synthetic dataset to test and tune the model and compares different experimental setups like random, kNN, and SVD+kNN recommendations. The goal is to increase business metrics like revenue, conversion rate, and average order value through effective product recommendations.
Branch And Bound and Beam Search Feature Selection Algorithms
Branch And Bound and Beam Search Feature Selection AlgorithmsChamin Nalinda Loku Gam Hewage This document discusses feature selection algorithms, specifically branch and bound and beam search algorithms. It provides an overview of feature selection, discusses the fundamentals and objectives of feature selection. It then goes into more detail about how branch and bound works, including pseudocode, a flowchart, and an example. It also discusses beam search and compares branch and bound to other algorithms. In summary, it thoroughly explains branch and bound and beam search algorithms for performing feature selection on datasets.
Amazon Item-to-Item Recommendations
Amazon Item-to-Item RecommendationsRoger Chen The document discusses recommendation algorithms used by Amazon, including traditional collaborative filtering, cluster models, and search-based methods. It focuses on Amazon's item-to-item collaborative filtering algorithm. This algorithm builds a similar-items table offline by finding items customers tend to purchase together. It then scales well to large data sets, provides high quality recommendations even with limited user data, and performs recommendations quickly.
Recommender Systems (Machine Learning Summer School 2014 @ CMU)
Recommender Systems (Machine Learning Summer School 2014 @ CMU)Xavier Amatriain The document summarizes a presentation on recommender systems given by Xavier Amatriain. It begins with introductions to recommender systems and collaborative filtering. Traditional collaborative filtering approaches include user-based and item-based methods. User-based CF finds similar users to a target user and recommends items they liked. Item-based CF finds similar items to those a target user liked and predicts ratings. Both approaches address sparsity and scalability challenges with dimensionality reduction techniques.
Developing Digital Literacy: 5 Ps for online learning
Developing Digital Literacy: 5 Ps for online learningRhona Sharpe The document outlines strategies for enhancing digital literacy in online learners through five key principles: prioritization, personalization, participation, presentation, and progress. It provides numerous actionable activities for educators, emphasizing the importance of learner engagement and tailored learning experiences. Successful online learners demonstrate self-regulation, collaboration, and effective technology use, with recommendations for further professional development in teaching and learning.
Solving the AL Chicken-and-Egg Corpus and Model Problem
Solving the AL Chicken-and-Egg Corpus and Model ProblemDain Kaplan Active learning (AL) is often used in corpus construction (CC) for selecting “informative” documents for annotation. This is ideal for focusing annotation efforts, but has the limitation that it is carried out in a closed-loop manner, selecting points that will improve an existing model. When there is no model, or the task(s) is even under-defined (such as studying corpora-less phenomena), use of traditional AL is inapplicable. To remedy this, we propose a novel method for model-free AL that focuses on utilising phenomena as desir- able characteristics. We introduce a tool, MOVE, that helps iteratively visualise and refine these characteristics. We show its potential on a real world case-study of a corpus we are developing.
Recommender.system.presentation.pjug.01.21.2014
Recommender.system.presentation.pjug.01.21.2014rpbrehm This document provides an introduction to recommender systems. It defines recommenders as information filtering systems that seek to predict a user's preferences for items. It describes different types of recommender systems, including content-based, collaborative filtering, and hybrid approaches. The history of recommenders is discussed from early manual filtering to modern automated approaches. Popular algorithms and open-source tools like Lenskit and Mahout are also introduced.
Multi Criteria Recommender Systems - Overview
Multi Criteria Recommender Systems - OverviewDavide Giannico The document discusses multi-criteria recommender systems (MCRS) and their role in addressing information overload through various techniques such as collaborative filtering and content-based approaches. It outlines the classification of MCRS, methodologies for preference modeling, and challenges such as managing intrusiveness and the development of new modeling approaches. Empirical results indicate that MCRS outperform single-rating techniques, demonstrating their effectiveness in improving recommendation precision.
Your own recommendation engine with neo4j and reco4php - DPC16
Your own recommendation engine with neo4j and reco4php - DPC16Christophe Willemsen The document discusses the creation of a recommendation engine using Neo4j and reco4php, addressing challenges like finding and serving relevant recommendations. It includes an overview of graph databases and technical challenges such as performance and simplicity. Additionally, it presents a practical demonstration of building a simple recommendation engine for GitHub repositories.
Summary of a Recommender Systems Survey paper
Summary of a Recommender Systems Survey paperChangsung Moon The document is a comprehensive survey on recommender systems (RS), covering foundational concepts, types of filtering (content-based, collaborative, hybrid), challenges like cold-start, and various similarity measures used in RS. It discusses modern trends, including location-aware systems and bio-inspired approaches, while addressing social filtering and research improvements in RS performance. Finally, it outlines future research directions, emphasizing the integration of diverse data sources and the evolution towards third-generation systems that leverage connected devices and user data.
Profile injection attack detection in recommender system
Profile injection attack detection in recommender systemASHISH PANNU The document discusses the detection of profile injection attacks in recommender systems, exploring the types of recommender systems, attack types, and detection attributes. It highlights the importance of recommender systems in e-commerce, lists various attack methods, and proposes an approach for detecting these attacks. Additionally, the performance of different models in identifying attacks is evaluated, with random forest, SVM, and neural networks showing better performance.
Recommendation Engine Project Presentation
Recommendation Engine Project Presentation19Divya The document discusses a recommendation system that analyzes user behavior to suggest relevant content and products based on their browsing patterns. It identifies problems with choice overload online and outlines a solution using a Java-based web tool for personalized recommendations. The document also details risk analysis, advantages, limitations, and future improvements for the system.
Recommender Systems and Active Learning
Recommender Systems and Active LearningDain Kaplan The document discusses recommender systems (RS) and active learning (AL), emphasizing their value for startups and established companies. It outlines RS objectives, methods, and addresses challenges like the 'cold start' problem for new users and items, while detailing how active learning can improve recommendation quality. Various AL strategies are compared for effectiveness in rating prediction accuracy, ranking quality, and the number of ratings acquired.
Online recommendations at scale using matrix factorisation
Online recommendations at scale using matrix factorisationMarcus Ljungblad The thesis presentation discusses the challenges of providing online movie recommendations at scale using matrix factorization, highlighting that 75% of daily movie starts come from recommendations. It outlines the performance of a system tested on 1 million items and emphasizes the need for balanced clusters and efficient data handling. The document concludes with a summary of setup and performance metrics, indicating a successful implementation with 74% effectiveness.
Requirements for Processing Datasets for Recommender Systems
Requirements for Processing Datasets for Recommender SystemsStoitsis Giannis This document summarizes the key requirements and challenges for processing datasets to be used in recommender systems based on three case studies. The main points are:
1) Recommender systems need to handle diverse social data from multiple sources in different formats and languages to support various recommendation scenarios and boost performance by combining data.
2) A common challenge is defining a metadata schema to transform and aggregate social data from different sources for federated recommender systems.
3) Case studies on a learning portal, open science platform, and multi-criteria rating dataset revealed additional challenges of data harvesting, anonymization, URI resolution, and developing algorithms that perform well with limited personalized data.
Building a Recommendation Engine - An example of a product recommendation engine
Building a Recommendation Engine - An example of a product recommendation engineNYC Predictive Analytics
Ad
Similar to [WI 2014]Context Recommendation Using Multi-label Classification (20)
Recommenders Systems
Recommenders SystemsTariq Hassan Recommender systems aim to recommend items like books, movies, or products to users based on their preferences. There are two main approaches: collaborative filtering, which recommends items liked by similar users, and content-based filtering, which recommends items similar to those a user has liked based on item attributes. Both have strengths and weaknesses, so hybrid systems combining the approaches can provide the best recommendations.
Overview of recommender system
Overview of recommender systemStanley Wang The document provides an overview of recommender systems. It discusses the typical architecture of recommender systems and describes three main types: collaborative filtering systems, content-based systems, and knowledge-based systems. It also covers paradigms like collaborative filtering, content-based, knowledge-based, and hybrid recommender systems. The document then focuses on collaborative filtering techniques like user-based nearest neighbor collaborative filtering and item-based collaborative filtering. It also discusses latent factor models, matrix factorization approaches, and context-based recommender systems.
PhD defense
PhD defense Giuseppe Ricci This document summarizes mathematical methods of tensor factorization applied to recommender systems. It discusses motivations and contributions, information overload and recommender systems, matrix and tensor factorization techniques in recommender system literature such as matrix factorization, singular value decomposition, high-order singular value decomposition, and parallel factor analysis. It also covers challenges in context-aware recommender systems and proposed solutions to incorporate contextual information.
Filtering content bbased crs
Filtering content bbased crsAravindharamanan S This document discusses various techniques for filtering and recommender systems, including content-based filtering, collaborative filtering, and hybrid approaches. It provides an overview of key concepts such as using user profiles and feedback to provide personalized recommendations. It also covers common recommendation algorithms like nearest neighbor collaborative filtering and discusses challenges like cold start problems and sparsity issues.
[AFEL] Neighborhood Troubles: On the Value of User Pre-Filtering To Speed Up ...
[AFEL] Neighborhood Troubles: On the Value of User Pre-Filtering To Speed Up ...Emanuel Lacić The document discusses the challenges and solutions in user-based collaborative filtering (ub-cf) for recommendations in the context of big data. It introduces a user pre-filtering technique to enhance runtime performance by optimizing the selection of similar users, thereby reducing the complexity of neighbor processing. Experimental results using a Foursquare dataset demonstrate that the adaptation can significantly improve recommendation speed and accuracy.
Recommender systems
Recommender systemsTamer Rezk The document discusses recommender systems, focusing on collaborative filtering and content-based approaches for personalizing recommendations across various domains, such as books and movies. It highlights the strengths and weaknesses of each method, detailing how machine learning is utilized to enhance user models, predict behaviors, and provide effective recommendations. Finally, it emphasizes the importance of combining both approaches to optimize recommendations and address challenges like cold start and sparsity.
Tag based recommender system
Tag based recommender systemKaren Li The document describes research on enhancing recommender systems through the use of user profiles and tagging systems. It discusses how user profiles can be used to provide personalized recommendations by describing a user's interests. It presents two research papers that studied how profile similarity and rating overlap between users can improve recommendation accuracy and user confidence. It also discusses how tagging systems can be leveraged by integrating user, tag, and resource dimensions. One paper proposes a personalized recommender model for folksonomies that extends the folksonomy by combining shared tags/resources and recommends tags and resources based on a user's profile and tagging history.
Effective Cross-Domain Collaborative Filtering using Temporal Domain – A Brie...
Effective Cross-Domain Collaborative Filtering using Temporal Domain – A Brie...IJMTST Journal The document surveys the effective use of cross-domain collaborative filtering (CDCF) in recommender systems, highlighting challenges such as user interest drift over time and data sparsity. It discusses methods to incorporate temporal domains to improve recommendation accuracy by considering current contextual parameters along with historical data. The paper proposes enhancements for modeling user interests by leveraging multiple time slices and contextual changes to provide more relevant recommendations.
Sergey Nikolenko and Anton Alekseev User Profiling in Text-Based Recommende...
Sergey Nikolenko and Anton Alekseev User Profiling in Text-Based Recommende...AIST This document summarizes a research paper on using distributed word representations and logistic regression to create user profiles for text-based recommender systems. The paper proposes clustering word embeddings to create document profiles, then using balanced logistic regression on individual users to assign weights to clusters for their profiles. This approach addresses issues of sparse data and uninformative clusters. An evaluation shows the regression-based recommender outperforms baselines on metrics like AUC, NDCG, and accuracy of top recommendations.
Mining Large Streams of User Data for PersonalizedRecommenda.docx
Mining Large Streams of User Data for PersonalizedRecommenda.docxARIV4 This document provides an overview of Netflix's approach to personalized recommendations. It discusses lessons learned from the Netflix Prize competition, including the effectiveness of matrix factorization and restricted Boltzmann machine models. It then describes Netflix's process for innovating recommendations called "Consumer Data Science", which uses A/B testing to evaluate new approaches. Finally, it outlines the key components of Netflix's personalized experience, noting that recommendations are integrated throughout the interface and account for differences within households.
factorization methods
factorization methodsShaina Raza The document summarizes five papers that address challenges in context-aware recommendation systems using factorization methods. Three key challenges are high dimensionality, data sparsity, and cold starts. The papers propose various algorithms using matrix factorization and tensor factorization to address these challenges. COT models each context as an operation on user-item pairs to reduce dimensionality. Another approach extracts latent contexts from sensor data using deep learning and matrix factorization. CSLIM extends the SLIM algorithm to incorporate contextual ratings. TAPER uses tensor factorization to integrate various contexts for expert recommendations. Finally, GFF provides a generalized factorization framework to handle different recommendation models. The document analyzes how well each paper meets the challenges.
An Example of Predictive Analytics: Building a Recommendation Engine Using Py...
An Example of Predictive Analytics: Building a Recommendation Engine Using Py...PyData This document discusses building a hybrid recommendation engine using Python to recommend Pubmed documents. It begins with an introduction to predictive analytics and recommender systems. Different types of recommender systems are described, including knowledge-based, content-based, collaborative filtering, and hybrid models. The document then outlines a hybrid model that performs content-based filtering on Pubmed documents using vector space modeling and weights documents, before applying collaborative filtering using the Python-recsys library to filter and recommend documents. Finally, it demonstrates the hybrid model on a Pubmed dataset and compares its performance to using Python-recsys alone.
The Network Research Project (NRP) expectations are to explain.docx
The Network Research Project (NRP) expectations are to explain.docxdennisa15 The Network Research Project (NRP) requires students to explore networking concepts in a focused manner, aiming for quality over quantity with a suggested length of 4-6 pages. Each project must include an introduction and conclusion, be well-formatted, and cite sources in APA 7th edition, with opportunities for draft submissions for feedback. Topics can include various network topologies, cloud concepts, network services, security implications, and more, with a focus on demonstrating understanding through research and analysis.
[UMAP 2016] User-Oriented Context Suggestion
[UMAP 2016] User-Oriented Context SuggestionYONG ZHENG This paper proposes a method called user-oriented context suggestion that suggests contexts to users based on their preferences. It aims to maximize user experience by recommending not just good items, but appropriate contexts for those items. Two algorithms are developed: one based on contextual rating deviations that identifies how a user's ratings change across contexts, and another that adapts techniques from item-oriented context suggestion. An evaluation on a music dataset finds the tensor factorization approach performs best, with the contextual rating deviations method also outperforming a simple baseline. Future work includes collecting better evaluation data and trying other contextual recommendation algorithms.
Notes on Recommender Systems pdf 2nd module
Notes on Recommender Systems pdf 2nd modulePRABHAKARK24 This document provides a comprehensive survey of recommender systems, primarily focusing on collaborative filtering algorithms and their impact beyond just rating prediction accuracy. It highlights the importance of factors such as diversity, novelty, unexpectedness, and their respective metrics, alongside examining the business value of recommender systems in terms of sales, customer loyalty, and consumer behavior. Key research directions and seminal papers in the field are also discussed to inform future advancements.
[Decisions2013@RecSys]The Role of Emotions in Context-aware Recommendation
[Decisions2013@RecSys]The Role of Emotions in Context-aware RecommendationYONG ZHENG The document discusses the role of emotions in context-aware recommender systems (CARS). It explores two classes of CARS algorithms: context-aware splitting approaches and differential context modeling. For context-aware splitting approaches, it examines which emotional contexts are most frequently used to split items or users. For differential context modeling, it analyzes which emotional dimensions are selected or weighted most highly for different algorithm components. The experimental results found that the emotions of end emotion and dominant emotion were the most influential across approaches. User splitting also generally outperformed item splitting.
Towards a Quality Assessment of Web Corpora for Language Technology Applications
Towards a Quality Assessment of Web Corpora for Language Technology ApplicationsMarina Santini The document outlines a two-step approach for creating and evaluating domain-specific web corpora for language technology applications, focusing primarily on medical terms. It discusses the extraction of term seeds from real-world use cases and scenarios, followed by the bootstrapping and assessment of the resulting corpora's quality using established metrics. The findings indicate that automatically extracted term seeds can yield corpora with similar domain-specificity as those created with hand-picked seeds, emphasizing the utility of quantitative evaluations in corpus quality assessment.
Advances In Collaborative Filtering
Advances In Collaborative FilteringScott Donald This document summarizes recent advances in collaborative filtering techniques for recommender systems. It describes how matrix factorization models have become popular for implementing collaborative filtering due to their accuracy. Neighborhood methods were also improved to be more accurate. The document outlines extensions that leverage temporal data and implicit feedback to further improve model accuracy. Key collaborative filtering approaches like matrix factorization, neighborhood methods, and techniques that combine their strengths are discussed.
SEppt
SEpptHemankita Perabathini The document describes an approach called Doc2Spec that infers resource specifications from natural language API documentation. It extracts method descriptions and class hierarchies from documentation using NLP. It builds action-resource pairs and infers automata specifications. An evaluation on 5 libraries found the inferred specifications have high precision and recall, and are useful for detecting real bugs in open source projects. Future work includes improving specification templates and applying it to other documentation.
Recommenders.ppt
Recommenders.pptNagendraBabu27244 The document discusses recommender systems and two main approaches: collaborative filtering and content-based recommending. Collaborative filtering recommends items based on preferences of similar users, while content-based recommending uses item attributes and a user's preferences on attributes to make recommendations. The document also describes LIBRA, a book recommender system that uses content-based filtering to learn user profiles from book ratings and provide recommendations.
Ad
More from YONG ZHENG (16)
[ADMA 2017] Identification of Grey Sheep Users By Histogram Intersection In R...
[ADMA 2017] Identification of Grey Sheep Users By Histogram Intersection In R...YONG ZHENG The document proposes a new approach to identify "grey sheep users" in recommender systems. Grey sheep users have unusual tastes and low correlations with other users. The approach represents each user as a histogram of their similarities to other users. It then uses outlier detection on the histograms to identify grey sheep users as the outliers with low similarities. The approach is tested on movie rating data and is shown to better identify grey sheep users compared to other methods. Future work involves applying this approach to other datasets and improving recommendations for identified grey sheep users.
[RIIT 2017] Identifying Grey Sheep Users By The Distribution of User Similari...
[RIIT 2017] Identifying Grey Sheep Users By The Distribution of User Similari...YONG ZHENG The document discusses a novel approach to identify 'grey sheep' users in collaborative filtering systems, which are users with unique tastes who do not align with broader user preferences. It outlines a four-step methodology leveraging user correlation distributions and outlier detection to improve user recommendations. Experimental results indicate the proposed method effectively identifies grey sheep users, with suggestions for future work and comparisons to existing techniques.
[WI 2017] Context Suggestion: Empirical Evaluations vs User Studies
[WI 2017] Context Suggestion: Empirical Evaluations vs User StudiesYONG ZHENG The document discusses context-aware recommender systems, focusing on context suggestions and their methodologies, including direct context prediction and indirect context suggestion. It presents empirical evaluations comparing these methodologies, highlighting the relevance of user contexts in enhancing recommendation quality. The study outlines experimental results and future work aimed at improving data collection and personalization in context prediction.
[WI 2017] Affective Prediction By Collaborative Chains In Movie Recommendation
[WI 2017] Affective Prediction By Collaborative Chains In Movie RecommendationYONG ZHENG This document discusses the exploration of affective prediction in movie recommendation systems, emphasizing the role of emotions and context-aware recommender systems. It presents methodologies for emotion acquisition and predictions, notably through independent and dependent collaborative approaches. The findings suggest that collaborative predictions enhance personalization quality, with future work aimed at expanding the evaluation to other domains and integrating context suggestion techniques.
[IUI 2017] Criteria Chains: A Novel Multi-Criteria Recommendation Approach
[IUI 2017] Criteria Chains: A Novel Multi-Criteria Recommendation ApproachYONG ZHENG This paper proposes a novel approach called Criteria Chains for multi-criteria recommender systems. Criteria Chains predicts ratings across multiple criteria in a chain, using previous predictions as context. It outperforms baselines by better utilizing relationships between criteria. The best method is to rank criteria by information gain to generate the chain, then use predicted criteria as context (CCC approach) to estimate the overall rating. Future work includes optimizing chain generation beyond information gain.
[EMPIRE 2016] Adapt to Emotional Reactions In Context-aware Personalization
[EMPIRE 2016] Adapt to Emotional Reactions In Context-aware PersonalizationYONG ZHENG This document discusses using emotions as context in recommender systems. It proposes two models that utilize emotional reactions data from movie ratings to improve context-aware recommender system algorithms. The models apply emotional regularization techniques to matrix factorization. One model regularizes based on similar emotional users, while another also considers original user similarities. Tests on a movie rating dataset show improvements over baselines, with emotional state during consumption more effective than after. Future work could explore emotional transitions over time.
Tutorial: Context In Recommender Systems
Tutorial: Context In Recommender SystemsYONG ZHENG This document provides an overview of a tutorial on context-aware recommender systems. The tutorial will cover traditional recommendation techniques, context-aware recommendation which incorporates additional contextual information such as time and location, and context suggestion. It includes an agenda with topics, background information on recommender systems and evaluation metrics, and descriptions of techniques for context-aware recommendation including context filtering and modeling.
Context-aware Recommendation: A Quick View
Context-aware Recommendation: A Quick ViewYONG ZHENG Context-aware recommendation systems take into account additional contextual information beyond just the user and item, such as time, location, and companion. There are three main approaches: contextual prefiltering splits items or users based on context; contextual modeling directly integrates context into models like matrix factorization; and CARSKit is an open source Java library for building context-aware recommender systems.
[UMAP 2015] Integrating Context Similarity with Sparse Linear Recommendation ...
[UMAP 2015] Integrating Context Similarity with Sparse Linear Recommendation ...YONG ZHENG This document summarizes a research paper on integrating context similarity with sparse linear recommendation models. It discusses contextual modeling approaches, including independent contextual modeling using tensor factorization and dependent contextual modeling using deviation-based and similarity-based approaches. It presents the sparse linear method (SLIM) and a contextual extension (CSLIM) that incorporates context similarity. Four methods for modeling context similarity - independent, latent, weighted Jaccard, and multidimensional - are described. Experimental evaluations on limited context-aware datasets are conducted to compare baseline algorithms like tensor factorization to the new similarity-based CSLIM approaches.
[IUI2015] A Revisit to The Identification of Contexts in Recommender Systems
[IUI2015] A Revisit to The Identification of Contexts in Recommender SystemsYONG ZHENG This document proposes a framework for identifying contexts in context-aware recommender systems (CARS). It defines contexts as any information that characterizes a user's situation. The framework models activities as having subjects (users), objects (items or other users), and actions (interactions). It provides three rules for context identification: 1) attributes of actions are contexts, 2) some dynamic attributes in user profiles can be contexts, and 3) some attributes of user objects can be contexts in social networks. The framework aims to clarify what should be considered contexts versus item content to improve CARS development and analysis.
[RecSys 2014] Deviation-Based and Similarity-Based Contextual SLIM Recommenda...
[RecSys 2014] Deviation-Based and Similarity-Based Contextual SLIM Recommenda...YONG ZHENG The document discusses the research of Yong Zheng on context-aware collaborative filtering and contextual slim (cslim) recommendation algorithms as part of his doctoral symposium at DePaul University. It covers the motivation, methodologies, and challenges associated with integrating contextual data into recommendation systems, highlighting both deviation-based and similarity-based approaches. The document also outlines ongoing work and future directions in improving effectiveness, scalability, and explainability of these algorithms.
[UMAP2013]Tutorial on Context-Aware User Modeling for Recommendation by Bamsh...
[UMAP2013]Tutorial on Context-Aware User Modeling for Recommendation by Bamsh...YONG ZHENG The document discusses context-aware user modeling for recommendations, highlighting the importance of understanding user interactions within different contexts, such as gifts versus personal interests. It explores the complexities of defining context, types of context (physical, social, interaction media, modal), and its implications for recommendation systems. Three architectural models for utilizing context—contextual pre-filtering, post-filtering, and modeling—are presented along with the challenges and methodologies for effectively integrating context into recommender systems.
Slope one recommender on hadoop
Slope one recommender on hadoopYONG ZHENG This document provides an overview of slope one recommender algorithms and their implementation in distributed systems using Hadoop and Mahout. It discusses slope one and weighted slope one recommenders, how they are implemented in Mahout, and how Mahout runs them in a distributed manner on Hadoop using mappers and reducers. It then describes experiments run on MovieLens data using this distributed slope one implementation and analyzes the results.
A manual for Ph.D dissertation
A manual for Ph.D dissertationYONG ZHENG The document provides an outline for a manual on writing a Ph.D. dissertation. It discusses introducing the dissertation, how to write and organize it, dissertation style, and good habits for writing a dissertation. Key sections include outlining the dissertation process and milestones, differences between papers/theses, common dissertation skeleton structures, principles for organizing sections, and tips for writing early and getting feedback.
Attention flow by tagging prediction
Attention flow by tagging predictionYONG ZHENG The document discusses the analysis of time-series data related to the usage of tags, specifically focusing on the tag 'bush' from 2004 to 2010. It details the data preparation, model fitting, and selection processes, leading to the conclusion that an ARIMA (1,1,1) model is effective for forecasting usage trends. Future research directions include using similar analysis for other political figures and topics to detect trends and seasonal effects.
[HetRec2011@RecSys]Experience Discovery: Hybrid Recommendation of Student Act...
[HetRec2011@RecSys]Experience Discovery: Hybrid Recommendation of Student Act...YONG ZHENG The document discusses a project aimed at improving participation of underprivileged youth in educational activities through a hybrid recommendation system that utilizes social network data. Several research questions are addressed, focusing on how to effectively recommend activities based on past enrollment, social dynamics, and the needs of both students and service providers. Initial experiments with data from two schools indicate promising results, suggesting a need for further research and larger datasets.
Recently uploaded (20)
Modern multi-proposer consensus implementations
Modern multi-proposer consensus implementationsFrançois Garillot Multi-proposer consensus protocols let multiple validators propose blocks in parallel, breaking the single-leader throughput bottleneck of classic designs. Yet the modern multi-proposer consensus implementation has grown a lot since HotStuff. THisworkshop will explore the implementation details of recent advances – DAG-based approaches like Narwhal and Sui’s Mysticeti – and reveal how implementation details translate to real-world performance gains. We focus on the nitty-gritty: how network communication patterns and data handling affect throughput and latency. New techniques such as Turbine-like block propagation (inspired by Solana’s erasure-coded broadcast) and lazy push gossip broadcasting dramatically cut communication overhead. These optimizations aren’t just theoretical – they enable modern blockchains to process over 100,000 transactions per second with finality in mere milliseconds redefining what is possible in decentralized systems.
ElysiumPro Company Profile 2025-2026.pdf
ElysiumPro Company Profile 2025-2026.pdfinfo751436 Description
ElysiumPro | IEEE Final Year Projects | Best Internship Training | Inplant Training in Madurai
Best Final Year project training center
Address:
First Floor, A Block, 'Elysium Campus, 229, Church Rd, Vaigai Colony, Madurai, Tamil Nadu 625020
Plus Code:
W4CX+56 Madurai, Tamil Nadu
+91 9944793398
[email protected]
Elysium Group of Companies established ElysiumPro in 2001. Since its inception, it has been the most sought-after destination for final year project development and research papers among the students. Our commitment to providing quality project training & documentation to students has always been exceptional. We deliver the final year engineering projects and technical documents that provide extra edge and industry exposure to land prestigious jobs and reputed institutions for higher studies. Students from all over the country avail of our services for their final year projects. On average, we develop 5000+ projects and research papers per year on varied advanced domains. Python, JAVA, PHP, Android, Matlab, LabView, VLSI, SIMULINK, Power electronics, Power System, Antenna, Machine Learning, Deep Learning, Data Science, Artificial Intelligence, data Mining, Big Data, Cloud Computing, IoT,
Hours of Operation: -
Sunday 10am-1pm
Monday 7.30am-8pm
Tuesday 7.30am-8pm
Wednesday 7.30am-8pm
Thursday 7.30am-8pm
Friday 7.30am-8pm
Saturday 7.30am-8pm
Web Site:
https://elysiumpro.in/
*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*
Youtube Geotagged Video:
https://youtu.be/QULY6XfuMyo
*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*
Slideshow Images (Google Photos):
https://photos.app.goo.gl/hVwQJtkeptA1JZKd9
*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*
GBP Listing:
https://goo.gl/maps/6d6hko6TsDYyeDrz9
*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*
Serving Areas:
https://www.google.com/maps/d/edit?mid=1-fsZogBiEAcjGP_aDyI0UKKIcwVUWfo&usp=sharing
*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*
Google Site:
https://elysiumpro-project-center.business.site
*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*
Google Sheet: https://docs.google.com/spreadsheets/d/1uXA07zxrUx2FCnBZWH80PpBZQrrX-2q1UBBe_0k3Yeo
*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*
Google Document: https://docs.google.com/document/d/1BU4ZHW_41XJm2lvTq9pWYUpZILAEmF9dWEw7-DBbWoE
*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*
Google Slides: https://docs.google.com/presentation/d/1uF8q6ueJWcAnhKTQsZxLE0Bo9PwgRNwCeuGV_ZgbSyU
*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*
Machine Learning - Classification Algorithms
Machine Learning - Classification Algorithmsresming1 This covers traditional machine learning algorithms for classification. It includes Support vector machines, decision trees, Naive Bayes classifier , neural networks, etc.
It also discusses about model evaluation and selection. It discusses ID3 and C4.5 algorithms. It also describes k-nearest neighbor classifer.
Introduction to Natural Language Processing - Stages in NLP Pipeline, Challen...
Introduction to Natural Language Processing - Stages in NLP Pipeline, Challen...resming1 Lecture delivered in 2021. This gives an introduction to Natural Language Processing. It describes the use cases of NLP in daily life. It discusses the stages in NLP Pipeline. It highlights the challenges involved covering the different levels of ambiguity that could arise. It also gives a brief note on the present scenario with the latest language models, tools and frameworks/libraries for NLP.
Learning – Types of Machine Learning – Supervised Learning – Unsupervised UNI...
Learning – Types of Machine Learning – Supervised Learning – Unsupervised UNI...23Q95A6706 Learning – Types of Machine Learning – Supervised Learning – Unsupervised Learning- semi supervised learning - The Brain and the Neuron – Design a Learning System – Perspectives and Issues in Machine Learning – Concept Learning Task – Concept Learning as Search – Finding a Maximally Specific Hypothesis – Version Spaces and the Candidate Elimination Algorithm
Tesla-Stock-Analysis-and-Forecast.pptx (1).pptx
Tesla-Stock-Analysis-and-Forecast.pptx (1).pptxmoonsony54 this is data science ppt for tesla stock (linear regression)
How to Un-Obsolete Your Legacy Keypad Design
How to Un-Obsolete Your Legacy Keypad DesignEpec Engineered Technologies For any number of circumstances, obsolescence risk is ever present in the electronics industry. This is especially true for human-to-machine interface hardware, such as keypads, touchscreens, front panels, bezels, etc. This industry is known for its high mix and low-volume builds, critical design requirements, and high costs to requalify hardware. Because of these reasons, many programs will face end-of-life challenges both at the component level as well as at the supplier level.
Redesigns and qualifications can take months or even years, so proactively managing this risk is the best way to deter this. If an LED is obsolete or a switch vendor has gone out of business, there are options to proceed.
In this webinar, we cover options to redesign and reverse engineer legacy keypad and touchscreen designs.
For more information on our HMI solutions, visit https://www.epectec.com/user-interfaces.
May 2025: Top 10 Read Articles in Data Mining & Knowledge Management Process
May 2025: Top 10 Read Articles in Data Mining & Knowledge Management ProcessIJDKP Data mining and knowledge discovery in databases have been attracting a significant amount of research, industry, and media attention of late. There is an urgent need for a new generation of computational theories and tools to assist researchers in extracting useful information from the rapidly growing volumes of digital data.
This Journal provides a forum for researchers who address this issue and to present their work in a peer-reviewed open access forum. Authors are solicited to contribute to the Journal by submitting articles that illustrate research results, projects, surveying works and industrial experiences that describe significant advances in the following areas, but are not limited to these topics only.
Fundamentals of Digital Design_Class_12th April.pptx
Fundamentals of Digital Design_Class_12th April.pptxdrdebarshi1993 Boolean Algebra and Combinational Logic Circuit
Deep Learning for Natural Language Processing_FDP on 16 June 2025 MITS.pptx
Deep Learning for Natural Language Processing_FDP on 16 June 2025 MITS.pptxresming1 This gives an introduction to how NLP has evolved from the time of World War II till this date through the advances in approaches, architectures and word representations. From rule based approaches, it advanced to statistical approaches. from traditional machine learning algorithms it advanced to deep neural network architectures. Deep neural architectures include recurrent neural networks, long short term memory, gated recurrent units, seq2seq models, encoder decoder models, transformer architecture, upto large language models and vision language models which are multimodal in nature.
Decoding Kotlin - Your Guide to Solving the Mysterious in Kotlin - Devoxx PL ...
Decoding Kotlin - Your Guide to Solving the Mysterious in Kotlin - Devoxx PL ...João Esperancinha Kotlin can be very handy and easy to use. Kotlin offers the possibility to develop code that is easy to understand, safe, immutable, and thus predictable and follows standards that avoid side effects. I realized that very quickly after I started my Kotlin journey that already amounts to more than 5 years.
This is the third version of this presentation focused on more detail explaining inline, crossinline, tailrec and as a bonus a quick run through unnamed classes.
[WI 2014]Context Recommendation Using Multi-label Classification
- 1. Context Recommendation Using Multi-label Classification Yong Zheng, Bamshad Mobasher, Robin Burke Center for Web Intelligence, DePaul University, Chicago IEEE/WIC/ACM Conference on Web Intelligence Aug 14, Warsaw, Poland
- 2. Intro – Recommender Systems • Information Overload Problem IR and RS • Recommender systems (RS) are the systems being able to provide recommendations to the end users. Currently, RS are popular everywhere: • E-Commerce: Amazon, Ebay, Newegg, etc • Social networks: Twitter, Facebook, etc • Movie/Stream: Netflix, Movie Pilot, Youtube, etc • Music: Pandora, Last.FM, etc
- 3. Intro – Type of Recommendations • [Item Recommendations] • [User Recommendations]
- 4. Intro – Context-aware Recommender • Context-aware Recommender Systems (CARS) CARS is a new type of RS which provide recommendations by adapting to users’ contextual situations. • Traditional RS: Users × Items Ratings • Contextual RS: Users × Items × Contexts Ratings • Assumptions behind: 1). User may have different preferences in different contexts; 2). Contexts are important in decision-makings. However, what CARS recommended are still items or users. Companion
- 5. New Application: Context Recommender • In this paper, we propose a new application: context recommender (CR), which is able to recommend or suggest appropriate contexts for users to select or consume items. • Sample of contexts: time, location, companion, etc Context Rec User RecItem Rec
- 6. Possible Examples (Amazon.com) Note: They are not REAL applications/examples right now.
- 8. Related Work L. Baltrunas, et al. "Best usage context predictions for music tracks", The 2nd Workshop on Context-aware Recommender Systems, ACM RecSys, 2010 This is the only work related to context recommendation, where the authors tried to provide suggestions of appropriate contexts to listen to music tracks. Pros: they proposed three KNN-based classifiers to suggest appropriate contexts; Cons: they proposed a specific application, but not a general delineation of the problem of context recommendations.
- 9. Contributions: Context Recommender • We formally provide the definition of CR • We propose the formal framework of CR applications • We discuss the algorithmic paradigms for CR • We examine the algorithms using multi-label classifications
- 10. Context Recommender 1. Definitions Context recommenders are the systems being able to recommend or suggest appropriate contexts for users to select or consume items. Examples: When is the best season to travel to Poland for user Tom? Who is the suggested companion to see “Titanic” with Tom? This book is better to be gifted to Mom or Kids? 2. Research Problems How to infer the appropriate contexts? And those contexts should be personalized or not? Example: The best season to Poland is always the same for all the users, or it could be personalized for different specific users? E.g. Tom likes winter much more than summer, where most users prefer summers in Poland.
- 11. Context Recommender 3. Context Recommendations App • There could be many other types, for example, it could be a group of users or items, instead of a single user or item. E.g. what is the best season to Poland by this group of travelers (e.g. Tour Group, etc), where the suggested contexts should meet the requirement of the group of users, instead of a single user. • In this paper, we focus on the general form: {User, Item} Contexts; a pair of <user, item> as input Input Output App {User, Item} Contexts The best season to Poland for Tom {User} Contexts The best travel season for Tom {Item} Contexts The best travel season to Poland
- 12. Context Recommender 3. Algorithmic Paradigms What are the possible algorithms to recommend contexts? After analysis, we propose two series of frameworks: 1). Direct Context Prediction We infer suggested contexts by <user, item, preferences>; In other words, contexts are predicted based on users’ previous preference histories associated with <item, contexts> 2). Indirect Context Recommendation We reuse the context-aware recommendation algorithms: In CARS, Therefore, we vary the choices of contexts, and finally recommend the contexts which can contribute to the best ratings user will give to the item.
- 13. Context Recommender 3. Algorithmic Paradigms Which algorithms can be applied in each category? 1). Direct Context Prediction Classification algorithms are the popular ones which are applied to this category, where they have been adopted in context predictions in the pervasive computing area. 2). Indirect Context Recommendation We reuse the context-aware recommendation algorithms; Therefore, all available CARS algorithms can be applied to. However, the drawback of this category is the high computational cost if there are too many contextual conditions in the data. And it also relies on how the CARS algorithms perform. In this paper, we focus on Direct Context Prediction using multi-label classification algorithms.
- 14. Context Recommender 4. Why Multi-label Classifications (MLC)? 1).Binary Classification (is this an apple?) 2).Multi-class Classification (is it an apple|orange|pear?) 3).Multi-label Classification (<round, apple, fruit, Mac>) In other words, MLC allows the system to select more than 1 labels from the set. Classification is used for context predictions; and MLC just fits the requirement of context recommendation task.
- 15. Context Recommender 4. Why Multi-label Classifications (MLC)? Two series of MLC algorithms: 1). Transformation Algorithms They can use traditional classification algorithms (e.g. decision trees, SVM, etc), and they transform the MLC task to multiple binary or multi-class classification tasks. So they do not need to develop new algorithms. E.g., binary relevance (BR), label powerset (LP), classifier chains (CC), k-labelsets (RAkEL) 2). Adaptation Algorithms Develop new classification algorithms to adapt to the MLC task. E.g., Binary relevance KNN (BRKNN), Multi-label KNN (MLKNN)
- 16. Context Recommender 5. Experiments and Evaluations (Algorithms) Toolkit: Mulan (MLC toolkit) and Weka Java-based library MLC algorithms: BR, LP, CC, RAkEL, MLKNN, BRKNN; Classification methods used in BR, LP, CC and RAkEL: KNNclassifier (KNN), decision trees (J48), naive bayes (NB),Bayesian nets (BN) and support vector machine (SMO) Baseline algorithms: 1).the three KNN classifiers: RatingBased (RB), BestContextVectorBased (BCVB) and BestContextBased (BCB) proposed by L. Baltrunas, et al 2).Non-personalized methods: such as most popular algs Data # of users # of items # of ratings # of labels Rating scale AdomMovie 69 176 1,010 8 1-13 LDOS 113 1186 2,094 17 1-5 TripAdvisor 2731 2269 14,175 5 1-5
- 17. Context Recommender 5. Experiments and Evaluations (Metrics) Inputs: User, Item, Binary Preference (Good or Bad) Outputs: A list of predicted contexts Time = Weekend Time = Weekday Companion = Kids Companion = Parents Companion = Girlfriend Real 0 1 0 0 1 Prediction 1 0 0 0 1 Y is the set of TRUE labels in the ground truth, and Z is the set of predicted TRUE labels. m = # of examples. Another metric is hamming loss which measures the average percentage of incorrectly predicted labels.
- 18. Context Recommender 6. Experimental Results Due to limited time, we only present results on LDOS data.
- 19. Context Recommender 5. Experimental Results (findings) 1).Personalization is required, because personalized algorithms work much better than the non-personalized ones. i.e., simply recommending most popular ones (e.g. most popular season people visiting Poland) is not enough; 2).LP algorithm using SMO as classifier is the best MLC algorithms among all data sets and all methods examined. They beat all the other algorithms. KNN-based approaches worked bad, because context-aware data are usually sparse. 3).SMO is the best classifier used for MLC algorithms, but it increases computation costs if data is large and there are many contextual labels. The alternative choice is Bayesian Nets which worked good and not time-consuming 4). About running performance: LP using SMO is the best choice, but both LP and SMO increase computation costs if data is large and there are many contextual labels. Solutions: a). Reduce the number of labels by pre-selections; b). Choose LP using Bayesian Nets
- 20. Conclusions and Future Work • We formally introduce and discuss the application and research problem of context recommendations (CR). We believe that context recommenders will provide many more novel applications and new recommendation opportunities for both practical use and the research community. • We propose the formal framework of CR applications and discuss the algorithmic paradigms for CR. • We examine the algorithms using multi-label classifications, and infer some significant findings and patterns as introduced previously. • Future work: Examine more other algorithms, and develop new evaluation metrics for this domain.
- 21. References • Context Recommendations [1].Baltrunas, Linas, Marius Kaminskas, Francesco Ricci, Lior Rokach, Bracha Shapira, and Karl-Heinz Luke. "Best usage context prediction for music tracks." In Proceedings of the 2nd Workshop on Context Aware Recommender Systems. 2010. [2].Yong Zheng, Bamshad Mobasher, Robin Burke. "Context Recommendation Using Multi-label Classification". In Proceedings of the 13th IEEE/WIC/ACM International Conference on Web Intelligence, 2014 • Multi-label Classifications [1].Tsoumakas, Grigorios, and Ioannis Katakis. "Multi-label classification: An overview." International Journal of Data Warehousing and Mining (IJDWM) 3, no. 3 (2007): 1-13. [2].Tsoumakas, Grigorios, Ioannis Katakis, and Ioannis Vlahavas. "Mining multi-label data." In Data mining and knowledge discovery handbook, pp. 667-685. Springer US, 2010.
- 22. Thanks! Yong Zheng, Bamshad Mobasher, Robin Burke Center for Web Intelligence, DePaul University, Chicago IEEE/WIC/ACM Conference on Web Intelligence Aug 14, Warsaw, Poland