Although I started this project as an experimental weekend thingy (to play around with Google App Engine), the project has shaped up well. Before you surf over to another blog, wondering what the hell I’m talking about, let me introduce you to “Personalized ARTICLE” aggregator (read as PARTICLE). The aim is to personalize a users online reading (just like what Findory did). Findory was an excellent service and I’ll be glad if I can achieve even an iota of what Greg created. This project is at very rudimetary and experimental stage. Rather than tapping into the users reading history on the site (monitored by the links clicked), the idea is to study how a users *interests*, scattered around at various “databases of interest” like, could be used to personalize online reading (news articles, blogs and more). This way the user could merrily browse the world wide web, bookmarking pages, doing his usual stuff and let PARTICLE worry about making this data useful.

Click here to try PARTICLE

Presently you need to provide PARTICLE with your username, which it uses to analyze your *interests* and present you with recent news stories you may like. It works well if you have a decent number of bookmarks in As I mentioned, the project is at a very rudimentary stage, so don’t feel disappointed by the results (ah! the unlucky few). I encourage you to play around with the app and recommend it to others to try. I’ll be making many changes/additions in the coming weeks.

Test drive PARTICLE at Kindly leave your feedback/comments/suggestions in the comments or send me an email at ‘anand at’.

[UPDATE] Yahoo! Research has a similar project called Garçon.

  • Aman

    Dude, Prof. Kristina Lerman at ISI, and Anon aur Rumi, her PhD students in my lab are working on very very similar things. May be you would be interested in looking at their work or I can get you talking with them, if u will.

  • Anand Kishore

    That’ll be great. Their suggestions and pointers would be very valuable. If you can, try to show them this app and get their feedback.

  • Abhimanyu Lad

    Works great dude! Some of the suggestions were really nice.
    Some way to organize the results would be useful (may be a tag cloud?), so that I can see what articles were suggested to me because of my interest in python, due to my interest in economics, and so on. Otherwise the resultset looks like a mixture of multiple themes.

  • Anand Kishore

    Oh! Good to hear that. I’ve come across people who weren’t satisfied by the recommendations (they have a very small set of bookmarks on The recommendations are great when you have a large dataset on

    I’m working on the link organization (maybe some clustering to remove duplicates). The tag cloud idea is great – i’ll have to think about it.