Scribd
Upload
2 views

Sentiment Analysis

Uploaded by

someswarichanamala
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
2 views

Sentiment Analysis

Uploaded by

someswarichanamala
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 14

SENTIMENT

ANALYSIS
Presented By:
Chanamala Bhuvaneswari
NRI Institute of Technology and Science
Department of Computer Science and Engineering
OUTLINE:
 Problem Statement(Should not include solution )
 Proposed Systems/Solution
 System Development Approach(Technology Used)
 Algorithm & Deployment
 Result
 Conclusion
 Future Scope
 References
PROBLEM STATEMENT

 Develop a sentiment analysis model to classify review as


positive or negative . Preprocess the review text using
techniques such as lower casting . Removing stop words and
lemmatization. Use the trained model accurately predict the
sentiment of new, Unseen reviews .
PROPOSED SOLUTION
Data Collection:
 Collect data from various sources available such as review
websites , social media , food blogs and online forums.
 Ensure the data is labeled with sentiment categories (e.g.,
positive, negative, neutral).This can be done manually or
through semi-supervised techniques.
Data Preprocessing:
 Remove common words that don’t contribute to sentiment.
 Split reviews into individual words or sentences.
 Convert text into lowercase and reduce words to their root
form
 Eliminate HTML tags , special characters and URL’s.
Deployment:
 Use cloud services(e.g., AWS , Google Cloud).
 Creating an API endpoint that can receive and preprocess new
review data.
 Update API using Flask , a lightweight web framework.
Evaluation:
 Retrain the model periodically with new data to maintain
accuracy and reliability.
 Use the deployed API to predict the sentiment of new, unseen
reviews and ensure it works as expected.
 Evaluate models using accuracy, precision, recall,F1 score.
SYSTEM APPROACH:
System Requirements:
 Processor: Multi-core processor (e.g., Intel i5/i7, AMD Ryzen
5/7).
 Windows, macOS, or Linux (Linux is often preferred for better
performance and compatibility with many machine learning tools)
 RAM: Minimum 8 GB (16 GB or more recommended for
handling large datasets)
 Storage: SSD with at least 256 GB of storage (512 GB or
more recommended)
 Programming Languages :Python (most commonly used for
sentiment analysis due to its extensive libraries and
community support)
Library Required To Build The
Model:

 Pandas
 NumPy
 NLTK
 Matplotlib
 Counter
 Sklearn
ALGORITHM & DEPLOYMENT
Algorithm Selection:
 Naïve Bayes algorithm is a popular choice for sentiment analysis for text classification
tasks where we need to determine the sentiment (positive, negative, neutral) of a
piece of text.
 Simple ,Fast , easy to implement and computationally inexpensive.
 It often achieves good results in Sentiment analysis tasks.
 Suitable for scenarios where quick implementation and good performance are
desired.

Data Input:
 Multinomial Naïve Bayes algorithm is preferred to use on data that is multinomially
distributed.
 It is one of the standard algorithms which is used in text categorization classification.
 Converting text into a numerical matrix where each row represents a document and
each column represents a feature (word, n-gram).
Training Process:
 Converting text into a numerical matrix where each row represents
a document and each column represents a feature (word, n-gram).
 Lowercasing: Converting all characters to lowercase.
 Removing Punctuation: Eliminating punctuation marks.
 Removing Stop Words: Removing common words that do not
carry significant meaning (e.g., "and", "the").
 Stemming or Lemmatization: Reducing words to their
base or root form.
Prediction Process:
 The new text data needs to be preprocessed same as the training
data.
 This includes tokenization, lowercase conversion , removing
stopwords and ensuring consistency in format.
RESULT:
 Here it indicates
Frequency of specific
words in reviews.
 In the graph the words
taken are good , great,
amazing , bad. Here good
and great have more
frequency than amazing
and bad.
 Accuracy of the tested data
is 80% .
 Hence, the overall result of
the tested data is positive.
CONCLUSION
 The primary goal is to understand the sentiment conveyed in a
piece of text, which can be positive, negative, or neutral.
 Sentiment analysis of food reviews is a valuable application of
natural language processing (NLP) techniques aimed at
understanding customer sentiments towards food products and
dining experiences.
 Various machine learning algorithms are applied to classify
sentiment. These may include logistic regression, naive Bayes.
FUTURE SCOPE
 Moving beyond overall sentiment, future research will focus on identifying
sentiments related to specific aspects or features within a text (e.g.,
sentiments towards different aspects of a product or service).
 There will be a push towards more nuanced sentiment analysis,
distinguishing between degrees of sentiment intensity (e.g., weakly
positive, strongly negative) to provide more granular insights.
 Integration of multiple modalities such as text, images, videos, and audio
for sentiment analysis to capture richer context and enhance accuracy.
 Incorporating sentiment analysis into AI-powered virtual assistants and
chatbots to enhance natural language understanding and improve user
interactions.
 Customizing sentiment analysis models to specific domains (e.g.,
healthcare, finance, politics) to improve accuracy and relevance of
sentiment predictions.
REFERENCES:
 Hamzah , A.A., M.F.J.J.o.U.S.S. Shamsudin, and Technology, Why
customer satisfaction is important to business? Journal of
Undergraduate Social Science and Technology, 2020.
 Alam , T.M. and M.J. Awan , Domain analysis of information
extraction techniques. International Journal of Multidisciplinary
Sciences and Engineering, 2018. 9: p. 1-9.
 Rehman, A.A., M.J. Awan, and I. Butt, Comparison and Evaluation
of Information Retrieval Models. VFAST Transactions on Software
Engineering, 2018. 6(1): p. 7-14.
 Hajjaji, Y., et al., Big data and IoT-based applications in smart
environments: A systematic review. 2021. 39: p. 100318.
 Sagiroglu, S. and D. Sinanc. Big data: A review. in 2013
international conference on collaboration technologies and systems
(CTS). 2013. IEEE.
THANK YOU

You might also like