12/8/2015 0 Comments Interest TrackerAt the time I was interning at The New York Times R&D Lab there was a big focus on semantic listening. Projects were focussed on capturing what a group of people are interested in or what is trending on internet. There are many use cases and implications of how NYTimes can use this to serve its readers. For instance, if NYTimes can know what its reader is discussing on a table it can be better suited to present him/her the articles or knowledge source to the person. I tried to organize information being shared among a group of people, for example NYTimes R&D group. I managed to direct their browser behavior (i.e. all the sites that they visited) through my application. Once I had the web urls I was able to fetch the content from the webpage and run LDA to find out the topics in it. Output was a histogram of topic weights with top scored words overlapped on it. Interestingly enough labs interest during lunch hours was directed towards the topics of food while other times it was governed mainly by technology blogs, news and a lot of programming helps. If this project be integrated with Listening table the interest of the people and direction can be evaluated and checked from time to time more accurately.
0 Comments
Leave a Reply. |
|