The Semantic Search Web UI is one of the KIM Platform front ends that allows you to access the powerful features of KIM through the Web. It is adapted to the IST World portal and supports queering based on both implicit and elicit relations among the data.

The main search functionality provided by all the available tools is organized among two main concepts:

• Entity - the objects in our knowledge base, from which the main four are: Organizations, Projects, Experts and Publications
• Documents - contextual description of the entities – e.g. the text of a Publication, description of a Project or Organization, etc.

Make your search effective, enjoyable, and easy enough so that you may have some time for THAT cup of tea or coffee once and then throughout your workday.

Explore entites IST World knowladge base!

If you look for an entity (i.e. Person or Organization) but you don't know its exact name, state the desired entity type and specify as much as you know about its name. Then you can search either for entities or right for documents related to any of them. E.g. "show me all organizations that are known as "Sirma"".

Explore the relations in IST World knowladge base!

In CORE one could ask for co-occurring entities, their rank or the deviations in their occurrence in time (timelines) conforming to restrictions like time period or a sub-set of the corpus. It allows incremental search and timelines based on extracted semantic information and on entities rank and co-occurrence. Even if the user interest is in relations that are not explicitly presented, still, one can easily find related entities and respective documents based on the associative relations found on the basis of co-occurrence in the same context (e.g. show me all Projects where Xerox and Imperial College co-occur ).

Let us suppose that we are interested in the relations of Hans Uszkoreit with other people. CORE allows asking about the most popular people that are mentioned together with Hans. The term “most popular” means “with the highest rank in the set of documents that also mentions Hans Uszkoreit”. The co-occurrence of the not so popular people will be considered as a less relevant part of the result set.

Since each document in the corpus has explicit date, we can track trends in time (timelines) for a set of entities based on their occurrence or co-occurrence frequency. Having these timelines, one can find out the important events in the life of a given entity following the extreme points of this function.