BEGIN:VCALENDAR
PRODID:-//hacksw/handcal//NONSGML v1.0//EN
METHOD:PUBLISH
BEGIN:VEVENT
DESCRIPTION:Click for Latest Location Information: http://semtech2011.semanticweb.com/sessionPop.cfm?confid=62&proposalid=4361\nOntotext?s Web Mining Framework (WMF) allows for extraction and normalization of specific types of information in the web. The process of acquiring of UK job announcements, for instance, involves:\n<ul>\n<li>focused web crawling, to acquire only the job-related pages from employer?s web sites\n<li>screen scraping, to harvest on regular basis large volumes of information from semi-structured  portals, e.g. job boards\n<li>text-mining and normalization to structure the information acquired from the web-sites\n<li>data merging and de-duplication, to compile a consistent, good quality knowledge base</li></ul>\nFurther, a typical web-mining project involves usage of pre-existing linked data datasets, databases, thesauri and other structured data (e.g. geographical information), which need to be transformed and integrated, before they can serve as a basis for the new dataset. The WMF was developed and matured in relation to few projects among which a UK online recruitment intelligence database and an aggregator of car offers collected from dealership web sites all over USA.  \nHere we present a recent project built on the WMF and OWLIM ? Pagane is aiming to organize all food related information in a knowledge base (KB). The company has started with the familiar realm of recipes, ingredients and techniques, but plans to expand the KB to include all aspects of procuring, preparing, consuming and enjoying food and beverages. The first application to be build on top of the KB is an enhanced recipe search, based on structured meta-data, such as calories per serving, fat content, sugar content, cooking time, etc. A user that is vegan and wants to cook a meal in under 30 minutes with beans and corn, will be able to quickly scan the web to come up with a few good recipes that match those criteria. At present the KB contains more than:\n<ul>\n<li>300,000 recipes from Epicurious, Food&Wine, Food Network and about 15 other portals\n<li>7,000 ingredients derived from SR23, LanguaL, Freebase and other datasets\n<li>350 cooking techniques</li></ul>
DTSTART:20110607T103000
SUMMARY:Baking with Data: from Jobs to Cars to Food
DTEND:20110607T111959
LOCATION: See Description
END:VEVENT
END:VCALENDAR