Topica was presented at the AIMashup challenge at ESWC2011 and was awarded the 2nd prize. The main idea of Topica is to profile a Location entity ( in this case a point of interest), based on the collective perception of it.

For a geographical bounded area, we extracted POIs using the Facebook Places and the Foursquare APIs. Using entity extraction services, including Open Calais, Zemanta, and DBpedia Spotlight we identified entity links (e.g. Person, Location, and Organisations), which we mapped to DBPedia.

We introduced the LinkedPOI Ontology, I haven’t made it available online yet. But should be posting about it soon.

More information can be found in this presentation, and the Topica website.

PreSense allows to express the streaming data as profiling feature of an entity. For example the data coming from mobile GPS is already used by Google for featuring your location, the data you get in your Iphone for tracking your heart bit is profiling one of your physiological features.

Have a look at the presentation