I am trying to find out if we should search and collect specific users based on race and gender and then code their tweets. OR if I collected a specific group of tweets based on a date range I could then figure out the users demographics based on the meta data.
Thanks for any input.
Hi Lauri,
If you know particular name of the user, then you can search and find him/her. After that, extract the required information (if available). Otherwise, you need to get the whole bunch of data (based on searching query) and find out the required information to be obtained. Here, the demographic knowledge might be not that much as it depends on tweets' nature, and your experience in analysing these tweets.
On the other hand, in regards to data collection features, it depends on the tools you are using (in case you're using Twitter Streaming API). Some tools provide the "Bio" of the user in each tweet and if you're lucky, you can find useful information there. Other tools don't provide such option.
HTH.
Samer
For location, Twitter allows user to add their location at their tweets.
For names, as Samer Sarsam added, you can know the names from region culture.
Thanks. Helpful. But I think, you can't tell just from an anonymous tweet gender and race, right?
Lauri,
AFAIK, this is not possible accurately without having such information located already in the tweet.
HTH.
Samer
You can probably extract or improve the gender/race information from the tweets' texts themselves.
You can take a look at Hillier, Analysing Real Texts: Research Studies in Modern English Language. Palgrave Macmillan. 2003 that identifies aspects of texts (literary narrative, newspaper reporting), spoken texts (women's talk, children's talk), written texts having spoken features (political speeches, fictional narrative in a regional dialect, television advertisements). You can take a look for the collection of different corpora and find out who works with them http://www.uclouvain.be/en-cecl-lcworld.html. You can also try Sally Planalp "Communicating Emotion" that draws conclusions on how texts are composed and who are their author http://catdir.loc.gov/catdir/samples/cam032/98049524.pdf.
Some information about racial behaviour could be extracted from scenarios of social interaction in "Social Interaction, globalization and computer-aided analysis." For approaches to classify textual information please consult "Opinion mining and lexical affect sensing".
http://catdir.loc.gov/catdir/samples/cam032/98049524.pdf
http://www.uclouvain.be/en-cecl-lcworld.html
https://he.palgrave.com/page/detail/analysing-real-texts-hilary-hillier/?sf1=barcode&st1=9780333584705
Thesis Opinion Mining and Lexical Affect Sensing
Book Social Interaction, Globalization and Computer-Aided Analysi...
- Badges
- Science topic
- Similar topics
- Political Science
- International History and Politics
- Globalization