07 June 2012

Another great meetup this month, with 2 talks and lots of good chats afterwards. 

Steve Butterworth (@stevewillbe) gave a really interesting talk on 50 different ways a computer can understand natural language and on each different type gave an indication as to which methods you should use to get the best results. His experience creating http://flumes.com/ meant that he has tried and tested a lot of the methods shown so it was really great to get some proper war stories from him too. 

Steve's slides can be found here: http://www.slideshare.net/humblefrog/50-ways-to-makes-sense-of-natural-language

I (@mrjaba) gave a slightly more "show and tell" presentation, wherein I showed off my latest pet project "Schlurp" which is a tool to assist with curating interesting news articles on my, rather empty, blog - mrjaba.posterous.com. It's a tool that uses the twitter streaming API and then extracts links from each tweets and counts any other tweet containing that link and keeps track of these. If the content is "interesting" then it will store it in Neo4j and visualize the output using sigma.js.

My slides can be found here:


Thanks again everyone for coming, looking forward already to next months!

blog comments powered by Disqus