Analysing Tweets Sentiments for Investment Decisions in the Stock Market

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

The increasing practice of using social media as the basis for decision-making has made social media an important alternative information source. This is, in particular, true for investors in the stock market due to their needs to gain dynamic, real-time information and strategic persons’ views. It is therefore very interesting to investigate the relationships between the sentiments of the text as published on social media and how they may influence investors’ minds. In this paper, we selected several influential Twitter accounts, inc. Bloomberg, Forbes, Reuters, WSJ and Donald Trump, for sentiment analysis using SentiStrength. We found a fair amount of agreement between the sentiments as generated by the tool and those assigned from investors’ point of view, esp. when plenty of positive words have been used in Tweets. However, we also discovered that not all Tweets with many positive words may generate positive sentiments in investors’ minds. Furthermore, we identified interesting differentiated sentiments expressed in different Tweeter accounts that may indicate the stance of their holders, e.g. using an upbeat tone thus to promote economic growth; or being conservative, thus maintaining one’s authority. Overall, we found many Tweets scored a neutral sentiment, as many of them contain references that their views cannot be determined without examining additional sources.

Original languageEnglish
Title of host publicationAgents and Multi-Agent Systems
Subtitle of host publicationTechnologies and Applications 2021
PublisherSpringer
Pages129-141
Number of pages13
ISBN (Electronic)9789811629945
ISBN (Print)9789811629938
DOIs
Publication statusPublished - 8 Jun 2021
Event15th International KES Conference on Agent and Multi-Agent Systems-Technologies and Applications 2021 - Virtual, Online
Duration: 14 Jun 202116 Jun 2021

Publication series

NameSmart Innovation, Systems and Technologies
Volume241
ISSN (Print)2190-3018
ISSN (Electronic)2190-3026

Conference

Conference15th International KES Conference on Agent and Multi-Agent Systems-Technologies and Applications 2021
Abbreviated titleKES-AMSTA 2021
CityVirtual, Online
Period14/06/2116/06/21

Keywords

  • Sentiment analysis
  • SentiStrength
  • Stock market price
  • Twitter

ASJC Scopus subject areas

  • Decision Sciences(all)
  • Computer Science(all)

Fingerprint

Dive into the research topics of 'Analysing Tweets Sentiments for Investment Decisions in the Stock Market'. Together they form a unique fingerprint.

Cite this