Intraday Stock Trend Prediction Using Sentiment Analysis.

Date of Submission

December 2018

Date of Award

Winter 12-12-2019

Institute Name (Publisher)

Indian Statistical Institute

Document Type

Master's Dissertation

Degree Name

Master of Technology

Subject Name

Computer Science

Department

Sampling and Official Statistics Unit (SOSU-Kolkata)

Supervisor

Mukherjee, Diganta (SOSU-Kolkata; ISI)

Abstract (Summary of the Work)

Traditionally, most of the studies involved in the prediction of stock price movement, using machine learning techniques, have utilized the historical stock price data and the technical indicator extracted from the data. However, such strategies do not use information about live news events that affect the stock prices. In this thesis, we focus on the publicly available tweets on twitter as a source of real time news. We extract the sentiment information from the tweets and use them (along with the stock data) to predict the intra-day price movement in the Indian stock market.We design a sentiment analyzer that identifies the sentiment of each tweet and sorts the tweet into one of three clusters (Trade, Feedback, and Miscellaneous). We modify our sentiment features using a cluster importance factor, a score that quantifies the importance of each clusters in the context of market importance, and use these modified features for price movement prediction via machine learning models. At the end of this thesis, we show that our trained models consistently outperform the random walk baseline accuracy of 33% for three-way stock trend classification and that the cluster importance factor improves the prediction accuracy of the random forest model.

Comments

ProQuest Collection ID: http://gateway.proquest.com/openurl?url_ver=Z39.88-2004&rft_val_fmt=info:ofi/fmt:kev:mtx:dissertation&res_dat=xri:pqm&rft_dat=xri:pqdiss:28843287

Control Number

ISI-DISS-2018-391

Creative Commons License

Creative Commons Attribution 4.0 International License
This work is licensed under a Creative Commons Attribution 4.0 International License.

DOI

http://dspace.isical.ac.in:8080/jspui/handle/10263/6957

This document is currently not available here.

Share

COinS