Freebase extractor has been retired in Wandora version 2015-08-06 because of the Freebase API deprecation.
Wandora's Freebase extractor makes an request to the Freebase API and converts the response to topics and associations. As with all other data in Wandora the extracted topics and associations may be modified, merged and visualized with the tools provided in Wandora.
The extractor is located in the dialog found in File > Extract > Topics > Freebase extractor...
Specifically the extractor uses the MQL Read Service of the API. A single extraction querys the API for a single Freebase topic and it's associated topics. The user can specify a one or more topics by their Freebase IDs or Freebase URLs in comma separated lists or by using the extractor's text search and selecting one or more topics matching the text query. The maximum amount of topics extracted and depth for a recursive query may also be specified. Note that the extractor may stop before the actual maximum amount of topics extracted is reached as extraction overlap in recursive queries may cause some topics to be counted multiple times.
Below an extraction is done by first creating a layer for the extraction and then searching for Freebase topics of Sibelius and picking the one representing the composer. The recursion depth is kept at 1 to extract only Sibelius' and it's immediate associations' topics.
The D3 Graph Visualization is used to visualize the extracted data. Written works, recordings and albums are shown to make up the bulk of data associated with Sibelius.
Next recursion is used to extract data of the Explorer, Voyager, Apollo and shuttle programs to visualize their shared associations. The graph below displays the interconnectivity between the programs and the Layer visualization shows a ratio of 0.33 for merged topics out of all topics in the layer structure.
Finally recursion of depth 3 is used to extract information from Hergé's series Adventures of Tintin and it's associated topics. The graph get's increasingly complex here and it mainly demonstrates large hub like topics that have large amount of topics associated with them and a distinct center of gravitation around which topics with many common associations are found.