Paper idea: make a better recommendation

For the final paper I would like to write something about the design of my team for the CHI student competition problem. It’s about helping the homeless people. We’re aiming at helping the 211 service specialist to make better recommendations on the social services to the homeless people, through a better designed computer recommendation application. (211 service is a nation-wide call-for-help service. People can call for help on: food, housing, transportation, financial assistance, etc. The specialist will recommend corresponding social service agencies to the clients.)

I am thinking of using structuralism approaches to analyze some existing recommendation systems (music recommendation, book recommendation, movie recommendation, etc), and maybe also some phenomenological concepts which can be applied here. Hopefully I can get some useful ideas about designing the recommendation system for the 211 service from this analysis.

My concern is, is it too broad to study recommendation systems in different areas? Would it be better to study only the existing book recommendation systems (or music,or movie)?


  1. Tyler Pace

    A very timely topic!

    There has been much research focused on recommendation systems in recent years. In fact, about a year ago Netflix announced that they hit the limit on the ability of their recommendation system and are offering a $1m reward to anyone who can improve their recommendation by some small %.

  2. yenning

    I think it is a very useful idea. My question is: how do you define if a recommendation system is good or bad? Will you use phenomenology or structuralism to analyze systems then find out how to design a recommendation system? For me, one possibility of your paper might be choosing one or two systems and analyze why you think it is good(or bad) by using the theories we discussed in class. Why users think they are good, how they use them, why designers design the system in this way. Or maybe you can choose one specific phenomenon(or design) in recommendation systems you are interested in and analyze it.

  3. jeffreybardzell

    I think this is a very good topic direction. Along with the suggestions offered by Tyler and Yenning, I would suggest narrowing it down further by picking one or at most two specific recommendation systems, and starting to concentrate just on them. When you do, it will be easier to answer Yenning’s question (what is a “good” recommendation system?) and also to narrow in on a specific research question that you can build your paper around.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s