Sentiment Analysis of English and Tamil Tweets using Path Length Similarity based Word Sense Disambiguation
Journal Title: IOSR Journals (IOSR Journal of Computer Engineering) - Year 2016, Vol 18, Issue 3
Abstract
Abstract: In social media, users have the privilege of connecting with people and extensively communicate, share information, discuss topics of recent trends. Friendster, LinkedIn, Instagram, Twitter are some media through which users can perform the activities as mentioned earlier. Twitter is a well-known microblog which allows user to express their opinion or sentiment in the form of tweets within maximum length of 140 characters. The sentiment of the user can be analyzed and interpreted using the concept called Sentiment Analysis (SA).Twitter is widely used in almost all parts of the world thus ensures the presence of multilingual tweets in expressing their sentiments. In this paper, Sentiment Analysis is employed to determine the polarity of English and Tamil tweets, which is therefore bilingual and it is further subjected to word sense disambiguation to figure out the contextual usage and further the sentiment of the words are classified using the Support Vector Machine which derives the polarity value of the words as positive, negative and neutral.
Authors and Affiliations
Kausikaa. N , V. Uma
Keywords
Related Articles
- EP ID EP159393
- DOI -
- Views 105
- Downloads 0
Integration of Search Interfaces in Deep Web Databases for Specific Domains
Abstract : The amount of information on Deep Web is growing rapidly, as most of the organizations put their content online behind the search forms. The information behind the search forms is often of very high quality an...
Analysis of Various Recommendation Systems
Abstract: In today’s world the only way to catch and keep user’s attention to one’s website is to provide them with the recommendation list to match the needs of user. In this paper the most popular and widely used recom...
A Proposed System for Tracking Invaded from the Crisis Zone Using RFID and Sensors
Abstract: This study is to show how sensors can help rescuing affected people from the casualty area caused by natural or human caused disaster. Sensors that will use Radio Frequency Identification to transmit the data t...
IVRS and DTMF based Voting System
Abstract: A Voting System will be developed where a person can vote using a cell phone. A phone connected tothe voting system, will decode the DTMF pulses sent by another phone(mobile/landline) belonging to the vot...
Skill Acquisition: An E-Learning Approach
The attention of the Nigerian Government for over twenty years now has been on youth’s empowerment through skill acquisition for self-realisation and economic independence. The Government and individuals, both in the pas...