Content Analysis is King

Emma Persky (Trampoline Systems)
Development
Location: New York West
Average rating: ***..
(3.20, 5 ratings)

Building content is only the first step to an interesting site, analysing that content to provide context can really makes it stand out. This type of context will allow your users to better understand that data they are looking at, and guide them into further exploring what your site has to offer.

For exampple simple techniques such as applying time-based discounting to content such ratings can immediately make scores far more relevant. This type of analysis is relevant to almost all statistics and data that contain a time based element, for example ratings on a review site, or date of a communication between users. Going further trend analysis can be used to let users see the historical performance of that content.

Another more advanced example is applying network analysis algorithms for exposing hidden relationships or finding cliques, clans or clubs in your data give your users a chance to explore the a much wider set of your content, and these techniques can help to not only push them in a direction they might be interested in, but also help them to understand how to pieces of content are related.

This talk is aimed at anyone (but mostly developers) with a vaguely mathematical background, who is interested in learning about some techniques you can use for analytics. They will be presented in an easy to understand and practical manner with references to existing tool sets wherever possible.

You’ll walk away with a “oh, that’s how they do that” kind of smile when you get when you see the beauty and simplicity in something you thought was complicated.

Emma Persky

Trampoline Systems

Emma Persky is currently a software engineer at Gumtree.com working on their location data and search project. Previously she could be found at Trampoline Systems building enterprise scale expertise search systems based on email processing and Organisational Network Analysis. She studied computing at Imperial College London where she completed her dissertation in Computer Vision Based Gesture Recognition Systems. She lies to describe her self as a passionate evangelist for the barcamp movement, an avid traveller and a vibrant storyteller. When not on a global expedition or planning her next barcamp, she can be found in London, England, working on one of her many unfinished data projects.

Lotus knows
Sponsors
  • IBM
  • Microsoft
  • Rackspace Hosting
  • 3DVIA
  • Authorize.Net
  • HP
  • Neustar, Inc.
  • OpenSRS
  • open{subnet}

Rob Koziura
(415) 947-6111
rkoziura@techweb.com

Kaitlin Pike
(415) 947-6306
kpike@techweb.com

View a complete list of Web 2.0 Expo contacts.