Knowledge is power and texts can contain all sorts of interesting information. But if there are large amounts of text, reading thru all that is time-consuming. Computers however can scan thru texts much faster and also the results always follow the same guidelines for finding the information while the humans may accidentally miss something or when there are multiple persons involved, they may use different criteria.
The results from information extraction can be:
- simple keywords, technical terms or names for search applications
- detailed information structures that tell who did what, where, how and when.
The webservice APIs available on this site include Named Entity Recognizer, an example of a tool to extracts names from the input text. There are also readily available software components that provide wider array of entities and their categories. The dependency parser needed for finding relations between words also exists. Please use the contact form to enquire about how to get access to these.