5

I'm currently querying Twitter to access trend data. I'm looking for an open alternative to source trend information. The main question is:

What's trending today?

I'd love to move away from using Twitter as the analytic tool and its very low request limit. Looking forward to the community's insights, thanks for any help.

I'm looking alternative streams of information about trends other than twitter. I'm very interested in sorting and categorizing trends by location. I'd prefer not to have to employ html scraping techniques.

If html scraping is the only options, which methods suit?

philshem
  • 17,647
  • 7
  • 68
  • 170
Gerard Downes
  • 151
  • 3
  • 1
    Not sure if you are asking for tools (like Topsy) that can help to analyze Twitter data or if you are looking for something else. Can you clarify? – Jeanne Holm Aug 06 '13 at 13:16
  • Topsy sounds interesting, but I was asking for other streams of trend information. I'll add that to the question. – Gerard Downes Aug 06 '13 at 14:27
  • Do you mean things to the effect of what's being googled or popular hashtags on instagram etc.? – batpigandme Aug 06 '13 at 15:58
  • Suggestions of different trend information streams would be helpful – Gerard Downes Aug 06 '13 at 16:12
  • 1
    Based on a blog post, it's fairly straightforward to collect historical and regional Google Search Trends data - http://opendata.stackexchange.com/a/4249/1511 – philshem Dec 31 '14 at 12:14

2 Answers2

4

Here's the best open data source for trend data I could think of: Wikipedia page view statistics. Derived datasets:

  • Wikitrends, a daily/weekly/monthly updated list of 10 most popular (as in page views) articles. Bonus: not only absolute, but also relative change top 10 lists (called uptrends/downtrends) are offered.
  • Wikpedia:5000, a weekly updated list of the 5000 most accessed articles.
  • Trending articles on Wikipedia finally seems pretty close to what you seek: popular articles in multiple language wikipedias during last 1, 3, 6, 12 or 24 hours. In my region of interest it shows the typical pattern of expected and surprising keywords.
ojdo
  • 2,804
  • 14
  • 31
  • These lists are very interesting, thanks. I'm very interested in sorting and categorizing trends by location. To pull information from these sources, I'm I mistaken for thinking that I'd need to use some html scraping techniques. – Gerard Downes Aug 08 '13 at 09:42
  • Indeed. Or try contacting the author; perhaps there is a convenient server backend which just isn't public. – ojdo Aug 09 '13 at 07:27
  • Interesting suggestion, thanks for the idea. All the best. – Gerard Downes Aug 09 '13 at 13:01
2

I've just been playing with Socrata's site metrics. You get a quite a lot of information, which is nice. Unfortunately, Socrata doesn't have very many users, and its users are probably quite strange compared to the rest of the world.

Avery Richardson
  • 271
  • 2
  • 2
  • 1
    Thanks for the input. Yeah it does seem to be a very specific type of trend data, I'll have to look deeper at it. Interesting. One a side note, FMS Symphony was inspiring. – Gerard Downes Aug 15 '13 at 16:00