Corpus Linguistics in the South #8: Voices from Below. Corpus Linguistics and Social Media
Saturday, 15 November 2014 at the University of Reading
I will be presenting a corpus linguistics talk with Alison Sealey on representations of animals via Twitter as part of the project, "'People,' 'Products,' 'Pests' and 'Pets': The Discursive Representation of Animals". The abstract for the talk is:
The Discursive Representation of Animals in the Twittersphere
As one strand of a funded research project that explores how people talk and write about a specific topic, we are constructing a corpus comprising texts from a wide range of genres. One of our sub-corpora will consist of tweets relevant to our topic, and this presentation will discuss the methods we are using to identify relevant tweets and process these for inclusion in our corpus. Thus the workshop theme most relevant to our talk is ‘methodological questions of mining and annotating data from online and social media platforms’.
We shall explain some of the approaches we have been using to capture relevant data, including filtering handles and hashtags, as well as tools such as Topsy, which sifts and retrieves tweets published since Twitter’s founding in 2006, and Keyhole, which is a social monitoring database that generates statistical summaries of the circulation and distribution of tweets.
Our topic is the discursive representation of animals*, and so we have had to find a method of identifying tweets that are about various kinds of animals, while eliminating those where terms denoting animals are used metaphorically (e.g. ‘yeah baby i am an ANIMAL in bed more specifically a koala I can sleep for 22 hours a day’).
We would like participants in the workshop to give us feedback on the work we present, and we welcome discussion about the issues raised, including the ethical and legal considerations applying to the use of this kind of data in linguistic research.
