Archive for freebase

Metaweb’s Freebase Now 60% Larger Than English Wikipedia – ReadWriteWeb

Metawebs Freebase Now 60% Larger Than English Wikipedia – ReadWriteWeb

Wow! that’s impressive. Freebase is really growing into a huge machine readable database of the world’s knowledge. Freebase is growing rapidly because bulk uploads are possible and scrapers are automatically feeding data into it. It will be interesting to see other serious applications besides Powerset making use of Freebase data. However, it is true that the interface is still complicated for the ordinary human users. Wikipedia is far simpler, user friendly and popular. Moreover, entering data into Freebase is not so flexible due to the strict data types required. Users may not always have perfect data to enter. That is why I intend to keep StYLiD interface simple and flexible for input. It is not at all intended to be another world’s database. Rather it is intended to be a simple platform to enable ordinary users to share a wide variety of structured data.

Advertisements

Comments (1)

Reconciliation in Freebase

Reconciliation : Freebase – The Worlds Database
A REST API to handle the reconciliation problem, summed up as “Tell me which freebase topic, if any, some data corresponds to.” I think reconciliation in Freebase is mainly duplicate detection. It would be nice if the system could automatically reconcile and merge multiple instances or types.

Leave a Comment

Freebase: Dispelling The Skepticism – ReadWriteWeb

Freebase: Dispelling The Skepticism – ReadWriteWeb

Freebase is definitely a nice and powerful application. It is different from Wikipedia though it is based on wikis because it maintains structured data. However, the Freebase interface seems to be quite overwhelming for the casual users. Wikipedia looks rather simple and easy to use. StYLiD is also based on similar idea as Freebase – enabling the users to create structured data. However, StYLiD is a lightweight application towards structured information sharing and not at all intended to be another world’s database (surely I cannot handle that). StYLiD is more like a blog for dynamic information sharing rather than a wiki. Moreover, StYLiD tries to consolidate multiple definitions by different users for the same type and produce richer and evolving conceptualizations. Freebase is cool and it will be interesting to see how people structure information. I am also trying to analyze user-defined types in Freebase and how they can be combined. Thanks to the Freebase API 😉

Leave a Comment

Freebase Wikipedia Extraction WEX

Freebase Wikipedia Extraction WEX

It is interesting that Freebase is making Wikipedia data available in machine readable XML form and promise to keep it updated. I hope it will be useful.

Leave a Comment