Actions
  • shareshare
  • link
  • cite
  • add
add
Other research product . 2022

Investigating the correlation between sentiment and stock-price fluctuations: Investigating the correlation between sentiment and stock-price fluctuations

Jørgensen, Oliver Lunding; Tækker, Tobias Lund; Paget, Marc David; Utzon, Bjørn Anton;
Open Access
English
Published: 01 Jan 2022
Publisher: Roskilde University
Country: Denmark
Abstract
This paper, seeks to examine the correlation between stock price and public sentiment expressed through social media. Through twitter scraping and pre- processing, sentiment can be extracted from text. The paper will be based on a heuristic approach to natural language processing. Furthermore, the paper will rely on the most common forms of sentiment analysis, using a rule-based and a machine-learning approach as a starting point and weigh these up against each other. Finally, we will continue with the best performing method, and weigh this up against real market data in a pursuit to find a correlation, should one exist. The paper found a sentiment-to-market accuracy 75%. And the accuracy score utilizing the rules-based approach of 72,72%.
Subjects by Vocabulary

EOSC: Twitter Data

Subjects

Natural Language Processing, NLP, Sentiment Analysis, Tweepy, space, web-scraping

Related Organizations
Related to Research communities
Digital Humanities and Cultural Heritage
moresidebar