Scribd
Upload
156 views

A Report of Recommender Systems: Ennan Zhai Peking University Zhaien@infosec - Pku.edu - CN

This document summarizes recommender systems. It begins by asking what recommender systems are, how they differ from search engines, and why they are needed. It then lists some famous conferences and researchers in the field. The document outlines existing recommender algorithms like content-based, collaborative filtering based on memory or models, and knowledge-based or hybrid approaches. Examples of recommender systems for e-commerce, music, movies and news are provided. The document compares advantages and disadvantages of content-based and collaborative filtering recommendations. Finally, it lists some famous recommender system datasets and poses questions for further discussion.

Uploaded by

Arockiaruby Ruby
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
156 views

A Report of Recommender Systems: Ennan Zhai Peking University Zhaien@infosec - Pku.edu - CN

This document summarizes recommender systems. It begins by asking what recommender systems are, how they differ from search engines, and why they are needed. It then lists some famous conferences and researchers in the field. The document outlines existing recommender algorithms like content-based, collaborative filtering based on memory or models, and knowledge-based or hybrid approaches. Examples of recommender systems for e-commerce, music, movies and news are provided. The document compares advantages and disadvantages of content-based and collaborative filtering recommendations. Finally, it lists some famous recommender system datasets and poses questions for further discussion.

Uploaded by

Arockiaruby Ruby
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
You are on page 1/ 19

A Report of Recommender Systems

Ennan Zhai Peking University [email protected]

First of all, we will answer the following questions:


What is the recommender systems?
What is the difference between recommender systems and search engine?

Why we need recommender systems?


What is the famous conferences in the area of recommender systems? What is the famous labs and people in the area of recommender systems?

Lets answer the former three questions here

Some Famous Conf: ACM Recommender Systems; WWW; SIGIR;


SIGCHI; KDD

Trans: Knowledge and Data Engineering;


Information System

Persons: J. Callan, R. White, P. Resnick, H. Yu and A.


Tuzhilin

Univ: CMU, NYU, Microsoft, NSU

Background

Which object to pick?

Recommender Systems !!!


E-Commerce: Amazon.com; eBay; Levis, Skieurope.com. Music: CDNOW; CoCoA; Music.Yahoo.com.

Movie: Nettfilx.com; Moviefinder.com.


News: Digg; GroupLens.

File: Credence.

The Existing Recommender Algorithms


We category the existing recommender algorithms into: Content-based Recommendation; Collaborative Filtering Recommendation; Heuristic-based or Memory-based; Model-based. Knowledge-based Recommendation; Hybrid Recommendation.

Content-based Recommendation

Users Feature

The Feature of Object Content

Content-based Recommendation
Zhang Y. Novelty and redundancy detection in adaptive filtering. In SIGIR02. Zhang Y. Maximum likelihood estimation for filtering thresholds. In SIGIR 01. Robertson S. Threshold setting and performance optimization in adaptive filtering. Information Retrieval, 2002.

Collaborative-based Recommendation
What is the collaborative-based recommendation? The Existing Collaborative-based Recommender System: Grundy System; Tapestry System; GroupLens; Ringo; PHOAKS System, and Jester System. Heuristic-based (Memory-based) and Model-based Algorithm

Collaborative-based Recommendation

Collaborative-based Recommendation
Default Voting;
Inverse User Frequency; Case Amplification; Weighted-majority Prediction.

Collaborative-based Recommendation
What is the model-based algorithm?

The others
What is the knowledge-based recommendation?

What is the hybrid recommendation?

Comparison of Typical Recommender Algorithms


Content-based Recommendation:
Results are intuitive and easy to interpret; No need for users access history info; Supporting by the mature tech. of classification learning;

Limited by the feature extraction methods; New user problem; Massive data and overhead.

Comparison of Typical Recommender Algorithms


Collaboration-based Recommendation:
No need for professional knowledge; Improving with the increasing of the user number; Easy to find users new interesting points;

Cold start problem; Limited by the history data; Easy to be attacked.

The Famous Data Sets for Recommender Systems


MovieLens;

EachMovies;
BookCrossing; Jester Joke; Netfix; Usenet Newsgroups.

Discussions
New recommendation algorithm?
Address some security problems (e.g., Sybil Attack and Shilling Attack)? Some Guarantees;

Thank you !!!

You might also like