Wikidata:WikiFactMine/What are we doing?
We are aiming to add numerous scientific statements to Wikidata. These statements are recovered by text mining from recent scientific papers, and so carry with them referencing and context: we also aim to add that information to Wikidata, as a structured citation.
Wikidata's content can be reused and displayed in numerous ways. That includes Wikipedia infoboxes, the rectangular standardised panels giving key facts in many articles.
The statements that WikiFactMine will be adding to Wikidata have simple forms: for example a given gene encodes a given protein; or a given drug is used to treat a given disease. Although the complexity of these statements is low compared to many to be found in the scientific literature, it is not at present possible to automate completely the process of extracting such data from a scientific paper.
We are, in effect, salami-slicing an unsolved problem. In software, we are finding likely candidate statements in papers: then with human assistance will decide when statements qualify for Wikidata. While this is not a novel idea in itself, we are providing innovative tools and approaches to make it practical. At same time, we have been consolidating some of the bibliographic support on Wikidata.