it’s WEET-IT Again , as mentioned in the previous post WEET-IT  is a Semantic knowledge engine  based on semantic web and open linked data that supports different use cases like Question answering , Keyword search , Comparisons and Relations our project achieved very promising results compared to the benchmarks of “Aqua log” and “Freya”  projects – the top places projects in the QALD-1 workshop depending on their evaluation datasets .

this demo demonstrates the first Question answering algorithm  of weet-it as well as the comparison  and Relation algorithms .

the answers are Full profiles in case of one answer like “what is the capital of Egypt ? ” the answer would be one entity which is Cairo
the full profile will contain the full abstract description of the entity and the high ranked properties for that entity , similar related entities grabbed from the data graph

some answers would be a mini profile  , this is the case when the answer would be one entity or two , maybe three let’s say it’s a short description with one or two of the most relevant property to that entity

finally in case of keyword search when answers number is large we prefer to preview the micro profile which contains only the picture and a little part of the abstract .

some results would be comparison tables in case of queries like ” compare between messi and beckham ”

or relation graphs previewing relations between two entities or more like : ” what is the relation between messi , ronaldinho , xavi”  the answer most probably would be that they are fooballers , played for barcelona and other relations found in the graph .

enough of talking ,, here’s the demo : please watch in HD

we are searching for funding for hosting on larger servers and expansions issues , for interesting organizations| persons  you can contact any of the team members  @hadyelsahar @Ali_Hosny  @HatemMira @sherifkandeel @oafifi @eslamsamymeg
as well feedback , offers & interaction is very welcomed

Advertisements