I am considering a project in which the content of the publication is complemented by relevant publicly available tweets from people in the area. But how could I programmatically find the appropriate tweets? I know that creating a structure that represents the meaning of natural language is pretty much the holy grail of NLP, but maybe there is some kind of tool that I can use, at least to reduce it?
Alternatively, I could just use hashtags. But this requires more work on behalf of users. I'm not very familiar with Twitter - do most people use hashtags (even for small scales) or rely on them to cut off a large data segment?
I will also be interested in capturing the status of Facebook (with permission to the poster, of course), and using the hashtag on Facebook is quite rare.
I could use a simple keyword search to roughly narrow the field, but more likely it would take human intervention to determine which tweets should be placed next to the content.
Ideas? Has this been done before?
source
share