Comments on: Data structure for Social news streams on Graph data bases

By: Typology Oberseminar talk and Speed up of retrieval by a factor of 1000

Typology Oberseminar talk and Speed up of retrieval by a factor of 1000 — Thu, 16 Aug 2012 11:55:08 +0000

[…] see from the slides we are planning to submit our results to SIGIR conference. So one year after my first blogpost on graphity which devoloped in a full paper for socialcom2012 (graphity blog post and blog post for source […]

By: Typology Oberseminar talk and Speed up of retrieval by a factor of 1000

Typology Oberseminar talk and Speed up of retrieval by a factor of 1000 — Thu, 16 Aug 2012 11:55:08 +0000

By: Rene

Rene — Mon, 13 Feb 2012 17:11:42 +0000

That would certainly work. The only problem is that this would not respect changes in the friendship graph. So if new friends come or go the list that you suggest would have to be updated.
Also in your case we would not really need a graph data base but could just use flat files (single index for every user) in order to store all the news updates.

By: Rene

Rene — Mon, 13 Feb 2012 17:11:42 +0000

By: Peter Ehrlich

Peter Ehrlich — Sun, 12 Feb 2012 21:59:18 +0000

Hello!
Thanks for writing this up, it’s very good to think about.
So this currently structures things as two dimensional list: one dimension traversing people, and the second the branches of their post. This means that to get a single list ordered by time, all branches must be queried and then merged.
What if one were to design a system slightly differently: Have each branch (of feed items) as it is now, but then rather than keeping a linked list of people for every follower, keeping a linked list of people’s feed items directly.
Thus when a post is made, it is added both to the top of its own branch, and to the top of of every follower’s feed.
Cheers
–Peter

By: Peter Ehrlich

Peter Ehrlich — Sun, 12 Feb 2012 21:59:18 +0000

By: Graphity: An efficient Graph Model for Retrieving the Top-k News Feeds for users in social networks

Sun, 05 Feb 2012 11:27:01 +0000

[…] Graphity works is explained in my first presentation and my poster which I both already talked about in an older blog post. But you can also watch this presentation to get an idea of it and learn about the evaluation […]

By: Graphity: An efficient Graph Model for Retrieving the Top-k News Feeds for users in social networks

Sun, 05 Feb 2012 11:27:01 +0000

By: 11 lessons learnt after my first scientific paper was submitted

11 lessons learnt after my first scientific paper was submitted — Wed, 02 Nov 2011 00:59:08 +0000

[…] become clearer when you really dig into it. It is amazing how all the practical runtimes of my graph data base index for social news feeds – let’s call it GRAPHITY – matched the theoretical runtimes. But while evaluating […]

By: 11 lessons learnt after my first scientific paper was submitted

11 lessons learnt after my first scientific paper was submitted — Wed, 02 Nov 2011 00:59:08 +0000