Browsing by Author "Singh, Vikas"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Sentiment Analysis on Political Speeches
    (Department of Computer Science & Engineering, NITS, 2022) Singh, Vikas
    Millions of human beings explicit their evaluations and mind approximately events, merchandise or businesses on online systems consisting of social networking web sites, blogs, and forums. The general view of people’s sentiment can be useful to make better choices, measure impact of positive choices or to slight content that can be dangerous for the general public however, the project is that the sentiment and views expressed by humans are in herbal Language, which isn’t understood with the aid of computers additionally, the quantum of records is so big that the mission cannot be completed manually. This paper indicates how natural Language Processing (NLP), and gadget learning (ML) may be used to make correct predictions of the underlying sentiment of quick texts such as a Tweet. We followed of the main techniques utilized in textual content sentiment extraction– Lexicon based totally technique and text classification technique however, we located the limitations of lexicon primarily based method for category of brief texts together with Tweets as a number of the Tweets did no longer include any lexicon within the dictionary. For the project of text classification, we have used gated recurrent unit(GRU) and linguistic-Neuro fuzzy system(L-NFS) algorithms on processed tweets and completed 62.0 percentile accuracy with 74.0 percentile in this paper, we also discuss the textual content pre-processing, tokenization and the restrictions of the strategies used.