Friday, September 10, 2010

Searching twitter for science related topics; best strategies? favorite searches?

I have been experimenting with different twitter searches relating to genomics, microbiology, evolution, and other work things and find it actually somewhat fascinating.  The key thing has been to try and find searches that pull out useful / interesting stuff but not crap.  For some topics, this is easy, like microbiome and metagenomics.  These work well for a few reasons I think including that they are technical but used somewhat commonly within the fields in which I am interested.

You can try to get a bit more specific with them if you look for the hashtag version of these words: #microbiome and #metagenomics.  This works OK for #microbiome but #metagenomics yields nothing right now.

Unfortunately, some potentially useful terms have been coopted by all sorts of non-interesting stuff. Consider evolution and bacteria, for example.  In these cases, the words are too commonly used either for non sciency stuff (e.g., evolution of skateboards just came up) or for non interesting stuff (e.g., kitchen cleaners come up a lot with the bacteria search).  In these cases, the hashtag versions are better though imperfect: #evolution and #bacteria.

The key seems to be to find terms that are used somewhat narrowly to be only about the topic you care about but are also used widely enough so that a lot of people in the field of interest use them.

Here are some of my current favorite/most useful search terms:

If you want, you can always go beyond these and do more advanced searches.  But I tend to stick to the simple ones much of the time.  I know many people do all their searching within twitter clients, and each has their own flavors of options, but for a few reasons I prefer to do searching directly at the twitter site.   


I continue to experiment with different approaches to this and would welcome any ideas for best ways to pull out useful material. 

Also, I am curious.  Do others out there who use twitter have some favorite terms they search for?

3 comments:

  1. Hey Jonathan, interesting post. It seams to me that the reason that common use terms are not useful is that context is lost. I wonder what would happen if the community posted OBO urls for common terms. Sort of like Latin organism names instead of common names. Only problem is twitter allows so few chars. ~ and unreadable.

    ReplyDelete
  2. I think that may be too complex. Perhaps we just need to invent new words ... though usually those ar pretty bad

    ReplyDelete
  3. Posting @MetaHIT tweets:

    1/ @phylogenomics here's how keywords searches look for us in @hootsuite: http://ow.ly/i/3Jp0 doubled w/ @trendistic charts http://ow.ly/2Cb7w

    2/ @phylogenomics but as you point out in your article, this is restricted to a rather easy field to follow

    3/ also like to use the google updates option. result for "genome evolution" in 2010, the peak is about Neanderthal Genome http://ow.ly/2Cbv0 (with a special search for "genome evolution" that you cite among your fav search!)

    Yohanan Winogradsky
    MetaHIT team

    ReplyDelete

Most recent post

Talk on Sequencing and Microbes ...

I recently gave a talk where I combined what are normally two distinct topics - the Evolution of DNA Sequencing, and the use of Sequencing t...