18. Engine Client e.g. portal, browser extension, rest API sentence splitter tokenizer (sentence into words) sentence parser fact generation (building semantic relations f r om parsed sentences) terms extraction semantic document representation Distributed on (1..n) servers Knowledge base storage semantic data model indexing text search engine Cluster of DB servers RDF model Front-end servers wrapper induction PDF Wrapper HTML Wrapper Custom Wrapper e.g. WIKIPEDIA Upload DOC Wrapper Text processing pipeline there are over 30 processors implemented in the grammar analyzer pipeline, switched on/off when needed cache
30. Heaven & Hell Wyjątkowy pomysł? Niekoniecznie. Odpowiedni moment wejścia na rynek? Bardziej. Znajomości? Przede wszystkim.
31. Winning Efforts Grand Prize Winner by 66% audience vote and 88% jury points “ Potentially huge—Jeff Clavier” “ Solving a very interesting problem” “ Nerdy winner of the night” Thought through very well” “ A ripe acquisition target”
32. :) “ Made for […] denser documents” “ Find a new friend in Topicmarks”
41. Bilans Organizing information on the cloud is growing 74% per year into a $5.5b market Cloud storage is exploding into a $33b market in 2015 Sources: OECD, Accenture, IDC IT Cloud Services Forecast, team analysis. Spending on organizing and retrieving research is worth $139b in 2015 CAGR 11% CAGR 58% Digitization of paper sources Multiple devices Remote collaboration Better backups Mainstreaming of research Information explosion Independent contracting Knowledge-based competitiveness CAGR 74% People organizing information through cloud storage will be a $5.5b market in 2015