Sentimental Technique Implementation on Textual Data
Baggam Sai Karthik, Dhavileswarapu NSSVV Durga Bharat, Md Afroz Alam, Bandari Tharun Sai, Mrinalini Rana, Smita Sharma, Suman Avdhesh Yadav · 2023
The increasing amount of real-time or near-real-time data in the big data age provides organizations with opportunities to make data-driven decisions. However, social media data, such as that generated by Twitter, is typically unstructured and difficult to manage. This paper proposes an efficient text data pre-processing approach and develops an algorithm for training decision tree, random forest, SVM, and logistic regression classifiers on a dataset. The algorithm weights the sentiment score based on the hashtag and cleaned text weights, and the most efficient data pre-processing strategy is identified using a novel technology called the hash vectorizer. The effectiveness of the algorithm is tested against manually sorted sentiments and sentiments created before text data pre-processing, and the resultsshow its accuracy is comparable to human-annotated feelings. This research contributes to the existing body of knowledge by providing a precise method for cleaning social media data and a new approach to sentiment analysis.