There’s a new twitter based application on the web that you don’t have to sign up to twitter to use:  Tweetolife.  The software is one of the outcomes of a study carried out by researchers at the Language, Interaction and Computation Laboratory of the University of Trento, who were working with data gathered for the Edinburgh Twitter Corpus, on which more background information is available to read as pdf.

Tweetolife can show you which genders use words more frequently (percentage statistics) and interestingly, can also show you at which times of day certain words peak or trough in frequency.  For example the word “class” peaks at just before 8.00am.

Of most use though is the gender differences “detailed query” which shows you which other words are most commonly used along with your search word – split by gender:  so for example, more men (@70%) use the word “replacement” than women (@30%) and with men it collocates most with “battery” while with women it’s “valve”.

Gender differences aside, it’s quite a useful mini-concordancer that could be used for collocation work with classes as it’s a bit more accessible than the Collins Co-build, which I’m not sure is still available anyway…  My only minor quibble is that there seem to be words that don’t register in the database – is this because people just don’t tweet them?  If so, it raises the interesting question of how and why we alter our language use when tweeting…..

Acknowledgement:  this was first spotted on Larry Ferlazzo’s site – thanks to Larry for that.

 

Correction:  when this was first posted, the initial paragraph read as follows:

There’s a new twitter based application on the web that you don’t have to sign up to twitter to use:  Tweetolife.  It’s the outcome – well, it’s one outcome of the development of Twitter corpus data by researchers at the University of Edinburgh, whose paper on the subject is available to read as pdf.

This has now been updated and corrected – thanks to Amaç Herdagdelen for providing the feedback on this!