- Create an account in order to vote and/or comment on the panels that interest you.
- Click on the specific panel title for more info and to comment. Select "search options" below for filtering options.
- Please note: The voting period is now closed.
Link to this page-
Collaborative Filters: The Evolution of Recommendation Engines
- Your vote:
- Yes No
- Organizer:
- Description:
- Recommendation engines have boomed in the era of social media. These panelists are experts at collaborative filtering systems. Citing Digg, Amazon, and Netflix as examples, they will have a high level discussion about the evolution of recommendation engines and how each approach is different.
- Questions
Answered: -
- What are the tradeoffs of publicizing algorithms? Do users care how recommenders work? Do they only care about results?
- How do user perceptions about recommendation logic effect the data they give to recommenders?
- How do privacy concerns enter into recommender designs?
- How do you deal with sparsity: not enough users per item, not enough movie renters per movie, etc?
- How do you jump-start recommendations? Early raters with no history, new items low statistics?
- How can we better serve "gray sheep", i.e. smaller sub-communities with unpopular views?
- What kinds of direct opposition, objection, annoyance, or complaints have you seen from users?
- Proposition: logic is overemphasized in collaborative filtering, user interface design and market messaging dominate results. True or false?
- How has the implementation of a recommendation engine affected the traffic and activity on sites?
- Is there a clear line between earnest user endorsements and cynical promotion? How might one account for such a distinction algorithmically?
- Panelists:
- Level:
- Advanced
- Category:
- Type:
- Panel
- Event:
- SXSW Interactive 2009
Developed for SXSW by Lindsey Simon