Sentiment analysis extracts the mood of a speaker or an author with respect to some subject or the overall contextual polarity of a document. In this paper, an algorithm is proposed for sentiment analysis of tweets extracted from social networking site, i.e., twitter. The comments of users are analyzed and are divided into two parts: positive and the negative sentiments. Algorithms used are keyword spotting and lexical affinity. Tweets are extracted from Twitter through REST API and real time. After that, filtration processes such as stemming, elimination of stop wordsetc are performed and then algorithms are applied on the remaining dataset. The correctness of those algorithms is checked using the results from the standard ALCHEMY API, and the accuracy of those algorithms is calculated individually. A new algorithm is proposed which is the hybrid of both keyword spotting and lexical affinity. The results of that algorithm are generated, and the accuracy is calculated and compared with rest of the algorithms. Machine learning is implemented using NLTK (natural language toolkit).
CITATION STYLE
Ruprah, T. S., & Trivedi, N. (2019). Sentiment analysis of tweets through data mining technique. In Advances in Intelligent Systems and Computing (Vol. 713, pp. 573–580). Springer Verlag. https://doi.org/10.1007/978-981-13-1708-8_53
Mendeley helps you to discover research relevant for your work.