3 ms·

Please elaborate, this sounds extremely interesting. Which firms are doing that with which data and what are the results one can get from that?

Robin_Message 16y ago
Timetric mines a variety of government and NGO data sources for economic data, including real time updates. I guess they would know who is collecting social data, if they aren't doing it themselves.
ig1 16y ago
Generally prop trading firms but also sector analysts at major investment banks. How they use this data is mostly considered propriety, but you can figure out the obvious uses. Using Google Insights and Amazon is actually immensely valuable is this space, as you can use product rankings/searches to figure out how products are doing well before any official announcements. For the social media stuff a lot of financial firms use third-party technology. There's a whole load of firms that do sentiment analysis on social networks. Most of the companies in this space build products for marketing/PR teams, but sector analysts in consumer markets often use them as well. Ironically the best way to track down the firms doing this stuff is to search Linkedin for people with data mining expertise working in the financial sector :-)
- dsplittgerber 16y ago
  Thanks for sharing your insights! Mind another question? How useful is that data really? I see the point in Amazon revenue #s of specific products etc, but "sentiment analysis".. really? Can one actually get actionable data from mining social networks like Twitter for sentiments? This reminds me of "long on sunny/short on rainy days"-trading patterns.
  - pierrefar 16y ago
    Earlier this year I was at a conference where the keynote talked about "sentiment analysis" of social media. Using real tweets as examples he went on to debunk the current techniques as useless, really, because they can't understand the sentiment. The current techniques he knew of and shared were all keyword based. The OP mentioned the problems with this kind of analysis, and I agree too. Edit: typo.
  - ig1 16y ago
    Well sentiment analysis for news seems to have gained mainstream acceptance, with major providers like Reuters offering apps in that space. For twitter, etc. it's still a relatively new field. There's a paper on it at http://arxiv.org/abs/1010.3003 http://arxiv.org/abs/1010.3003

Robin_Message 16y ago

Timetric mines a variety of government and NGO data sources for economic data, including real time updates. I guess they would know who is collecting social data, if they aren't doing it themselves.

ig1 16y ago

Generally prop trading firms but also sector analysts at major investment banks. How they use this data is mostly considered propriety, but you can figure out the obvious uses. Using Google Insights and Amazon is actually immensely valuable is this space, as you can use product rankings/searches to figure out how products are doing well before any official announcements. For the social media stuff a lot of financial firms use third-party technology. There's a whole load of firms that do sentiment analysis on social networks. Most of the companies in this space build products for marketing/PR teams, but sector analysts in consumer markets often use them as well. Ironically the best way to track down the firms doing this stuff is to search Linkedin for people with data mining expertise working in the financial sector :-)

dsplittgerber 16y ago
Thanks for sharing your insights! Mind another question? How useful is that data really? I see the point in Amazon revenue #s of specific products etc, but "sentiment analysis".. really? Can one actually get actionable data from mining social networks like Twitter for sentiments? This reminds me of "long on sunny/short on rainy days"-trading patterns.
- pierrefar 16y ago
  Earlier this year I was at a conference where the keynote talked about "sentiment analysis" of social media. Using real tweets as examples he went on to debunk the current techniques as useless, really, because they can't understand the sentiment. The current techniques he knew of and shared were all keyword based. The OP mentioned the problems with this kind of analysis, and I agree too. Edit: typo.
- ig1 16y ago
  Well sentiment analysis for news seems to have gained mainstream acceptance, with major providers like Reuters offering apps in that space. For twitter, etc. it's still a relatively new field. There's a paper on it at http://arxiv.org/abs/1010.3003 http://arxiv.org/abs/1010.3003