TRAP@NCI

Sentiment analysis of motorcycle manufacturers through twitter feeds

O'Connor, Colm (2014) Sentiment analysis of motorcycle manufacturers through twitter feeds. Diploma thesis, Dublin, National College of Ireland.

[img]
Preview
PDF (Diploma)
Download (1MB) | Preview

Abstract

The purpose of this project is to detect sentiment towards various motorcycle brands through tweets. I have chosen Twitter due to three factors, the ease, or supposed ease, of access to tweets, the information is in text form and limited to 140 characters per tweet, and finally as it is used for daily gossip, talk, brainstorming, general opinions and commentary.

Twitter’s new application programming interface (API) had brought up challenges in getting historical data, and as such any analysis could only be done on data generated from the time of connection to the stream. The data received from the stream was much more than 140 characters. Each tweet contained many variables, including a time stamp, unique tweet id, the user’s bio, and most importantly the language of the tweet as determined by Twitter. Code had to be written to cleanse the tweets and to anonymise the usernames within the tweets themselves.

Through the use of a sentiment lexicon the brands can be checked for positive and negative sentiment. Based on the methods used, the search terms, and the data presented we can infer which brands are more prevalent than others in the “twittersphere”, the collective postings made on Twitter.

From my analysis BMW appears to be the most popular brand with twice the mentions of that of its closest competitor Honda. BMW appears in 45% of the tweets, whereas Honda is in 22%. Although this seems to be a categorical reading that BMW is twice as popular as Honda, and one hundred times more popular than Harley, who comes in at 1%, this is not the case. On inspection of BMW’s data, its popularity appears to be mainly derived from its motorcar following and its sponsorship of events, most recently the BMW PGA Championship at Wentworth. England.

Item Type: Thesis (Diploma)
Subjects: Q Science > QA Mathematics > Electronic computers. Computer science
T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science
Divisions: School of Computing > Higher Diploma in Science in Data Analytics
Depositing User: CAOIMHE NI MHAICIN
Date Deposited: 12 Dec 2014 16:28
Last Modified: 12 Dec 2014 16:28
URI: http://trap.ncirl.ie/id/eprint/1851

Actions (login required)

View Item View Item