Page 52 - 2023

Ph.D.
(Computer Applications)
STOCK MOVEMENT PREDICTION USING COMPREHENSIVE
DATA MODEL

Ph.D. Scholar : Darji Dhara Natvarlal
Research Supervisor : Dr. Satyen M. Parikh

Regi. No.: 17276211003
Abstract :
The stock market is notorious for its volatility, nonlinear and dynamic nature. The
potential to forecast stock price movement is a critical topic for both researchers and
industry experts, but developing a model that can do so is challenging. A predictive
system that is able to forecast the direction of a stock price movement helps investors
make appropriate decisions, improves profitability, and hence decreases possible losses.
Forecasting of the stock market prices and their directional changes play an important
role in financial decision making.

In today's stock markets, investor sentiment is an important factor in a stock's future
worth. In this technological era and fast expansion of the internet, social media and
Internet of Things (IoT) services generate massive amounts of information per instant. It
is a challenging task to analyse this type of data because the volume of the data is large
and it also expands rapidly. It produces structured and unstructured data so sentiment
analysis is effective for processing text data and filtering the outcomes to provide
contextually relevant insights. The goal of this research is to develop a model using stock
price historical data, technical indicators and mood information in social media that could
be used to predict stock market fluctuations (up or down). The traditional approach is not
capable of processing such data in real time so many organizations and researchers
have tried to implement new big data frameworks in order to handle data for a long. The
various big data frameworks are used to handle such data due to its 5v characteristics of
volume, value, velocity, variety, and veracity.
The researchers are also working out for the same and give hints to carry out this work in
more efficient way in future for improvement in forecasting the stock trend movement.
This research work was carried out with the aim on implementing a strategy that
provides an acceptable level of accuracy in the prediction of stock market trends. For the
same, the comparative study was carried out with four targeted values to identify and
make the model more effective. The model deals with both the analytical parts which
require for stock market forecasting.

One of the analytical parts is sentiment analysis which focuses on gathering online
published financial market related news and Twitter tweets for BSE top 100 stock

27