International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 10 Issue: 07 | July 2023
www.irjet.net
p-ISSN: 2395-0072
Live Twitter Sentiment Analysis and Interactive Visualizations with PyLDAvis using Streamlit Sabbineni Lakshmi Gopi Koushik 1, Chinthapatla Navyasri2, Gurram Keerthana3, Chunduru Jahnavi4 ---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - Sentiment analysis, often known as opinion
posted by customers on social media. Amongst the most difficult things of sentiment analysis is appropriately recognizing the sentiment expressed in text, particularly true for ambiguous sentences in which the sentiment may be unclear. Researchers and practitioners in sentiment analysis are always attempting to increase the accuracy and reliability of sentiment analysis approaches by employing modern machine learning algorithms and natural language processing techniques. Irrespective of challenges, sentiment analysis has grown in importance in the age of big data.
mining, is a technique for determining the emotional tone or attitude indicated in a piece of text, such as a tweet, review, or news story. As Twitter has many influential users, it became an important tool for communication and information exchange on various platforms including politics, business, and entertainment. This paper provides a user-friendly online application for 'sentiment analysis of live Twitter data and interactive visualizations using pyLDAvis' built on Python's VADER module and the Streamlit framework. Our application asks users to enter a topic or phrase of interest and then uses the Twitter API to stream tweets in real-time. We use VADER to do sentiment analysis on incoming tweets and illustrate the results in real-time using various interactive charts and plots. Further, we implement PyLDAvis to do topic modeling on the tweets and display the topics and keywords connected with them. In a dynamic and interactive manner, our application allows users to explore the emotion and themes of Twitter conversations connected to their areas of interest.
1.1 Problem in Existing System The problem with the existing system is limited training data. To learn how to identify sentiment in text, sentiment analysis systems rely significantly on training data. However, if the training data is inadequate or polarised, the system may be unable to identify sentiment in new material effectively. Another problem of not having a display of sentiment analysis results is that it can be difficult for users to quickly and readily understand the sentiment of the text. Without a visual representation, consumers may have to go through huge amounts of text or data to understand the overall sentiment, which can be time-consuming and ineffective. Patterns and trends in sentiment analysis data may not always be seen from the text alone, but visualization can help. A visualization, for example, may illustrate the frequency of good, negative, and neutral sentiments across time, allowing viewers to observe changes in sentiment over time.
Key Words: Sentiment Analysis, pyLDAvis, Streamlit Framework, Vader library.
1.INTRODUCTION Natural Language Processing (NLP) is a rapidly emerging computer science topic that has grown in significance over the past few years. NLP is focused on teaching machines how to read and process human language in the same way that humans do. This entails creating algorithms and computer processes capable of analyzing, interpreting, and producing human language data such as text or speech. NLP offers a wide range of practical applications, including text classification and sentiment analysis, as well as machine translation and chatbots. Sentiment analysis, commonly referred to as opinion mining, is an important method for assessing and analyzing sentiment represented in text data. Sentiment analysis uses computer techniques and algorithms to recognize and classify the emotional tone or attitude indicated in a piece of text, such as a tweet, review, or news story. Sentiment analysis is a tool that can analyze client feedback and evaluations, helping businesses in understanding the benefits and drawbacks of their products or services. It can also assist customer care personnel in promptly identifying and responding to bad feedback or complaints
© 2023, IRJET
|
Impact Factor value: 8.226
2. PROPOSED WORK The proposed work provides an architecture for live Twitter sentiment analysis, which is based on a lexiconbased approach and includes classification issues in realtime. This model is entirely built on querying real-time tweets from the Snscrape and pre-processing them into a corpus of words using a Lexicon-based method, where the terms are specified.
|
ISO 9001:2008 Certified Journal
|
Page 1187