Today Extractiv gives you a little more control over your ability to manage your content by launching our new On-Demand platform. On-Demand complements Extractiv's existing Crawling platform. The Crawling platform provides users with a means to transform the output of large web crawls into structured content through batch jobs. By contrast, the On-Demand platform enables users to process their own documents and URLs one at a time.
On-Demand gives you two additional types of control that may be necessary for certain applications:
- Immediate responses to requests, which enables use in a live application.
- Ability to process user content which is not accessible through the Web (although URLs are also permissible as input).
Requests can be submitted to the REST API by 1) browser, 2) command-line, or 3) programmatically.
Processing a URL online can be as easy as providing a 'url' parameter to a web request. For example, a story on ABC news today about the hero Captain Jack Conroyd's crash landing can be processed using this link (output is in machine-friendly JSON format -- stay tuned later this week for a new human-friendly HTML output format). Key entities extracted from this document included John F. Kennedy Airport (AIRPORT_FACILITY), Atlantic Southeast Airlines (AIRLINE_ORG), the Hudson River (RIVER), and of course Jack Conroyd and Larkin Newby (PERSON).
Finally, the On-Demand platform is in the free beta stage -- just like the Crawling platform -- so give it a whirl! We've love to hear your feedback, so please leave comments.