Predicting UK stock market short-term activity and returns from Daily Mail Online

Loading...
Thumbnail Image

URL

Journal Title

Journal ISSN

Volume Title

School of Business | Bachelor's thesis

Date

2017

Major/Subject

Mcode

Degree programme

Rahoitus

Language

en

Pages

31

Series

Abstract

In this paper, I’m deriving a direct measure of investor sentiment from Daily Mail Online news articles using SentiWordNet lexical resource for opinion mining, and I find that it has the ability to predict London Stock Exchange market activity and returns. More precisely the results state that all news – not only financial news – can offer a significant input in predicting stock market activity and that the direction of the effects differs between news categories. The dataset used in this paper is unique; it consists of 1,139,243 Daily Mail Online news articles published between years 2008 and 2017.

Description

Thesis advisor

Lof, Matthijs

Keywords

opinion mining, finance, sentiment analysis, online news, investor sentiment, big data, computational linguistics, Daily Mail

Other note

Citation