Comments on: Who wants to build a “find a date using Twitter” service? http://www.unstructuredventures.com/uv/2009/01/21/who-wants-to-build-a-find-a-date-using-twitter-service/ Translating Business Strategies into Financial Models Thu, 22 Jul 2010 21:35:00 -0700 hourly 1 http://wordpress.org/?v=3.0.1 By: Marc Vermut http://www.unstructuredventures.com/uv/2009/01/21/who-wants-to-build-a-find-a-date-using-twitter-service/comment-page-3/#comment-2850 Marc Vermut Thu, 03 Sep 2009 01:20:00 +0000 http://www.unstructuredventures.com/uv/?p=697#comment-2850 For that, you should check in with Brynn Evans who uses Mechanical Turk a bunch for her research. For that, you should check in with Brynn Evans who uses Mechanical Turk a bunch for her research.

]]> By: Marc Vermut http://www.unstructuredventures.com/uv/2009/01/21/who-wants-to-build-a-find-a-date-using-twitter-service/comment-page-3/#comment-2628 Marc Vermut Wed, 02 Sep 2009 23:20:12 +0000 http://www.unstructuredventures.com/uv/?p=697#comment-2628 For that, you should check in with Brynn Evans who uses Mechanical Turk a bunch for her research. For that, you should check in with Brynn Evans who uses Mechanical Turk a bunch for her research.

]]> By: Taylor Davidson http://www.unstructuredventures.com/uv/2009/01/21/who-wants-to-build-a-find-a-date-using-twitter-service/comment-page-3/#comment-2627 Taylor Davidson Wed, 02 Sep 2009 15:26:37 +0000 http://www.unstructuredventures.com/uv/?p=697#comment-2627 I wonder how Mechanical Turk could crunch away the data sets ... I wonder how Mechanical Turk could crunch away the data sets …

]]> By: Taylor Davidson http://www.unstructuredventures.com/uv/2009/01/21/who-wants-to-build-a-find-a-date-using-twitter-service/comment-page-3/#comment-2626 Taylor Davidson Wed, 02 Sep 2009 15:24:22 +0000 http://www.unstructuredventures.com/uv/?p=697#comment-2626 I should have noted that @manukumar's tweet about Thread is what made me think about this again :)<br><br>The real question is which set of data is better, which set of data can be structured easier, and which one works better. How would a TweetPsych-inspired algorithim work on a similar set of data for the same FB users? How would the results from FB and Twitter differ?<br><br>Sadly, since there is / could be a large degree of variability between how people use Twitter and FB (and email), the results may vary differently even for a specific individual. I should have noted that @manukumar's tweet about Thread is what made me think about this again :)

The real question is which set of data is better, which set of data can be structured easier, and which one works better. How would a TweetPsych-inspired algorithim work on a similar set of data for the same FB users? How would the results from FB and Twitter differ?

Sadly, since there is / could be a large degree of variability between how people use Twitter and FB (and email), the results may vary differently even for a specific individual.

]]> By: Marc Vermut http://www.unstructuredventures.com/uv/2009/01/21/who-wants-to-build-a-find-a-date-using-twitter-service/comment-page-3/#comment-2625 Marc Vermut Wed, 02 Sep 2009 14:50:11 +0000 http://www.unstructuredventures.com/uv/?p=697#comment-2625 Yesterday, after reading about Thread (<a href="http://mashable.com/2009/09/01/thread/" rel="nofollow">http://mashable.com/2009/09/01/thread/</a>) it crossed my mind that Facebook has a more richer set of personality and social network data (what you like, what you've seen, who you know, how often you communicate with them, where you've been) via all of the applications. And then I saw this Twitter analytics article (<a href="http://mashable.com/2009/08/30/analyze-twitter-content/" rel="nofollow">http://mashable.com/2009/08/30/analyze-twitter-...</a>) and a description of TweetPsych.<br><br>TweetPsych (<a href="http://tweetpsych.com" rel="nofollow">http://tweetpsych.com</a>) was built by by Dan Zarrella, and:<br>"uses two linguistic analysis algorithms (RID and LIWC) to build a psychological profile of a person based on the content of their tweets. The service analyzes your last 1000 tweets and works best on users who have posted more than 1000 updates." Yesterday, after reading about Thread (http://mashable.com/2009/09/01/thread/) it crossed my mind that Facebook has a more richer set of personality and social network data (what you like, what you've seen, who you know, how often you communicate with them, where you've been) via all of the applications. And then I saw this Twitter analytics article (http://mashable.com/2009/08/30/analyze-twitter-...) and a description of TweetPsych.

TweetPsych (http://tweetpsych.com) was built by by Dan Zarrella, and:
“uses two linguistic analysis algorithms (RID and LIWC) to build a psychological profile of a person based on the content of their tweets. The service analyzes your last 1000 tweets and works best on users who have posted more than 1000 updates.”

]]> By: Taylor Davidson http://www.unstructuredventures.com/uv/2009/01/21/who-wants-to-build-a-find-a-date-using-twitter-service/comment-page-3/#comment-2624 Taylor Davidson Wed, 02 Sep 2009 08:57:25 +0000 http://www.unstructuredventures.com/uv/?p=697#comment-2624 (returning to an old conversation)<br><br>It's interesting, I spend a lot of time talking to people publicly using Twitter across the world, across cultural contexts, time contexts (my monday morning is someone's sunday night, or monday dinnertime), personal and professional contexts, etc...<br><br>bringing me to the notion of social dynamics and intimacy: intimacy is a shared dynamic, and it's only created when both people understand the context of the situation and the community; as we create new communities do we create new <a href="http://www.taylordavidson.com/writing/2009/05/19/ambient-intimacy/" rel="nofollow">standards for intimacy</a>? can individuals change as fast as societies? (or, isn't is obvious that hand-wringing and dismay over cultural changes are to be expected because groups of people [call them generations] can't change as fast as society [the sum of generations at any one point in time] ? )<br><br>more groups, more splintering, more opportunities for cultural translators (tools and people)... (returning to an old conversation)

It's interesting, I spend a lot of time talking to people publicly using Twitter across the world, across cultural contexts, time contexts (my monday morning is someone's sunday night, or monday dinnertime), personal and professional contexts, etc…

bringing me to the notion of social dynamics and intimacy: intimacy is a shared dynamic, and it's only created when both people understand the context of the situation and the community; as we create new communities do we create new standards for intimacy? can individuals change as fast as societies? (or, isn't is obvious that hand-wringing and dismay over cultural changes are to be expected because groups of people [call them generations] can't change as fast as society [the sum of generations at any one point in time] ? )

more groups, more splintering, more opportunities for cultural translators (tools and people)…

]]> By: Taylor Davidson http://www.unstructuredventures.com/uv/2009/01/21/who-wants-to-build-a-find-a-date-using-twitter-service/comment-page-3/#comment-2623 Taylor Davidson Wed, 02 Sep 2009 08:36:27 +0000 http://www.unstructuredventures.com/uv/?p=697#comment-2623 Thinking back, perhaps I've downplayed the difficulty in mining a tweetstream or a lifestream to truly understand people; I wonder, in your mind, what is the best heuristic publicly available online to create a proxy or an understanding of a person's complete self (online and offline)? Thinking back, perhaps I've downplayed the difficulty in mining a tweetstream or a lifestream to truly understand people; I wonder, in your mind, what is the best heuristic publicly available online to create a proxy or an understanding of a person's complete self (online and offline)?

]]> By: Taylor Davidson http://www.unstructuredventures.com/uv/2009/01/21/who-wants-to-build-a-find-a-date-using-twitter-service/comment-page-3/#comment-2593 Taylor Davidson Wed, 05 Aug 2009 16:22:20 +0000 http://www.unstructuredventures.com/uv/?p=697#comment-2593 Do you think Twitter's "reputation score" will be "right"?<br><br>link: <a href="http://siliconangle.com/ver2/2009/08/04/how-user-authority-and-reputation-will-change-twitter-for-the-better/" rel="nofollow">http://siliconangle.com/ver2/2009/08/04/how-use...</a> Do you think Twitter's “reputation score” will be “right”?

link: http://siliconangle.com/ver2/2009/08/04/how-use...

]]> By: Marc Vermut http://www.unstructuredventures.com/uv/2009/01/21/who-wants-to-build-a-find-a-date-using-twitter-service/comment-page-3/#comment-1537 Marc Vermut Thu, 19 Mar 2009 23:16:13 +0000 http://www.unstructuredventures.com/uv/?p=697#comment-1537 What's interesting about the idea is the distributed nature of interaction enabled by Twitter. You don't have to go to a separate website or email account. Notifications are pushed to you. Your personality is (theoretically) assessed via a sentiment analysis (<a href="http://en.wikipedia.org/wiki/Sentiment_analysis" rel="nofollow">http://en.wikipedia.org/wiki/Sentiment_analysis</a>) of your tweetstream (or lifestream, maybe a pet project for the Gnip team).<br><br>Huge mashup potential with psychological tests. There is already a tools that guess your Meyers-Brigg (<a href="http://www.typealyzer.com/" rel="nofollow">http://www.typealyzer.com/</a>) and age/sex based on blog writing.<br><br>That said, I think it is more potent for business connections, interest groups and flash invites to targeted events (tied to a mobile LBS). All of this, of course, is based upon the availability of real working software that delivers accurate text analysis. What's interesting about the idea is the distributed nature of interaction enabled by Twitter. You don't have to go to a separate website or email account. Notifications are pushed to you. Your personality is (theoretically) assessed via a sentiment analysis (http://en.wikipedia.org/wiki/Sentiment_analysis) of your tweetstream (or lifestream, maybe a pet project for the Gnip team).

Huge mashup potential with psychological tests. There is already a tools that guess your Meyers-Brigg (http://www.typealyzer.com/) and age/sex based on blog writing.

That said, I think it is more potent for business connections, interest groups and flash invites to targeted events (tied to a mobile LBS). All of this, of course, is based upon the availability of real working software that delivers accurate text analysis.

]]> By: Marc Vermut http://www.unstructuredventures.com/uv/2009/01/21/who-wants-to-build-a-find-a-date-using-twitter-service/comment-page-3/#comment-1987 Marc Vermut Thu, 19 Mar 2009 23:16:13 +0000 http://www.unstructuredventures.com/uv/?p=697#comment-1987 What's interesting about the idea is the distributed nature of interaction enabled by Twitter. You don't have to go to a separate website or email account. Notifications are pushed to you. Your personality is (theoretically) assessed via a sentiment analysis (<a href="http://en.wikipedia.org/wiki/Sentiment_analysis" rel="nofollow">http://en.wikipedia.org/wiki/Sentiment_analysis</a>) of your tweetstream (or lifestream, maybe a pet project for the Gnip team).<br><br>Huge mashup potential with psychological tests. There is already a tools that guess your Meyers-Brigg (<a href="http://www.typealyzer.com/" rel="nofollow">http://www.typealyzer.com/</a>) and age/sex based on blog writing.<br><br>That said, I think it is more potent for business connections, interest groups and flash invites to targeted events (tied to a mobile LBS). All of this, of course, is based upon the availability of real working software that delivers accurate text analysis. What's interesting about the idea is the distributed nature of interaction enabled by Twitter. You don't have to go to a separate website or email account. Notifications are pushed to you. Your personality is (theoretically) assessed via a sentiment analysis (http://en.wikipedia.org/wiki/Sentiment_analysis) of your tweetstream (or lifestream, maybe a pet project for the Gnip team).

Huge mashup potential with psychological tests. There is already a tools that guess your Meyers-Brigg (http://www.typealyzer.com/) and age/sex based on blog writing.

That said, I think it is more potent for business connections, interest groups and flash invites to targeted events (tied to a mobile LBS). All of this, of course, is based upon the availability of real working software that delivers accurate text analysis.

]]>