« More data beats better algorithm at predicting Google earnings | Main | Is Search Advertising a Giffen Good? »

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

Jeff

You guys have written some great article around data - are there any books you can recommend on the subject?

Thanks,
Jeff

Anand Rajaraman

@Jeff:

Would love to help you, but "data" is a very high-level term. Could you be more specific: databases, data mining, analytics, machine learning, etc? Or did you mean to suggest that I periodically write posts reviewing books on data-analysis related topics?

Srinivas

Yes, there are network effects, but I'm not sure they are so strong as in auctions or social networks for this to be a winner takes all game.

(1) Companies frequently need to move data between different applications and end up doing this through a whole bunch of ways (ETL tools for example). I think in many cases, the network bandwidth is not the bottleneck - usually the data transformations and load process itself tend to be a bottleneck. So, it maybe feasible for two different apps to be in two different data centers and consume each others data.

(2) There are a large number of different data centers/hosting services out there today. If the network effects were so strong, I'd have expected there to be far fewer of those.

Jeff

Sorry about the lack of specifics Anand - that was like me asking you to write about the ocean - maybe a better way to ask is how does the class you teach unfold, or what subjects do you cover in your class? If there a logical flow from databases to data mining, and so on?

Thanks,
Jeff

The comments to this entry are closed.