International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 09 Issue: 05 | May 2022
p-ISSN: 2395-0072
www.irjet.net
Social Media Mining: Sentiment Analysis on Twitter Data Rashi Bhattad, Mansi Satpute, Ujjwal Mishra Rashi Bhattad, PICT, Pune Mansi Satpute, PICT, Pune Ujjwal Mishra, PICT, Pune --------------------------------------------------------------------***--------------------------------------------------------------------three ways which are descriptive, predictive, and Abstract –
prescriptive. Many models are created and by using machine learning algorithms, the most valid and practical models are found. The final stage is to deploy the best models by making calculated decision made by humans, and operational decision, made by machines, which will answer the question. Lastly, with continuous monitoring and measuring of the models, the success of the models’ outcomes is evaluated. The four stages are altered slightly to progressively increase the effectiveness of the specific data mining application in question.
“Social Media Mining”, Which essentially means a sentiment analysis done on people on social by using various statistics and analyzing algorithms of the pattern of people’s activity on social media sites. We were able to get the data from various social media sites and the Machine Learning algorithms were implemented on them for keyword analysis. Which is basically analyzing the patterns about the general populations’ behavior regarding a specific topic via keyword searches, mentions or tweets done on the social media sites. We had to get the raw data from these social media sites and then process it in such a way that it was viable for performing the process of Machine Learning on it. We have made use of K-Nearest Neighbor (KNN) algorithm to train the Machine, as well as Natural Language Processing (NLP) for enabling the Machines to understand the Human language. And Natural Language ToolKit (NLKT) for sentiment analysis to obtain insight of the audience on social media.
1.2 Relevance Web-based Media Mining includes web-based media, network investigation, and information mining to give a helpful stage to understudies and task administrators to comprehend the extent of online media mining. It presents the issues emerging from web-based media information and presents essential ideas, impending problems, and pragmatic calculations for network investigation and information mining. With the assistance of the Machine Learning course, we can apply ideas, standards, and techniques in different situations of online media mining.
Key Words: Social Media Mining, Sentimental Analysis, Natural Language Processing.
1. INTRODUCTION
1.2
In today’s world social media sites such as Google, Facebook, YouTube, Twitter, etc. play an important role in every individual’s life. People on these sites upload and download data according to their needs. All these social media sites are filled with trillion bytes of data which can be recognized on different aspects of social media and human interactions. To better understand these interactions on the web, social media mining can be done to get a better understanding of the latest trends. 1.1
The scope and objectives of social media mining are to: Understand social aspects of the Web with Social Theories, Social media and Mining. Learn to collect, clean, and representable social media data. To measure essential properties of social media and simulate social media models. Find and analyze communities in social media. Understand how information propagates in social media. Understanding friendships in social media, performing recommendations, and analyzing behavior. Study or ask interesting research issues. Startup ideas/research challenges. Learn representative algorithms and tools
Literature Survey
The first stage is to find different data sources, geared around answering a specific question and figuring out a way to compile various forms of raw data from multiple sources for this data can be used as input for data mining. The next stage is to explore the data by using visualization tools to improve and filter the initial idea and to transform the data to answer our queries. Moving forward the next stage is to model the data by creating and applying multiple analytical algorithms that will find patterns and draw assumptions based on data presented. Data modelling can be done by
© 2022, IRJET
|
Impact Factor value: 7.529
Scopes and Objective
2. THEORETICAL DESCRIPTION The most common way of mining social information includes a mix of measurable Machine Learning, science, and statistics. The initial step is to accumulate and handle social information from various web-based media sources. Aside from online media stages like Twitter, or YouTube, information diggers likewise remove information from
|
ISO 9001:2008 Certified Journal
|
Page 797