US 9,811,518 B2
Systems, methods, and software for questionbased sentiment analysis and summarization
Jochen L. Leidner, Eagan, MN (US); Kingsley Martin, Beverly Shores, IN (US); Trace Liggett, Rosemount, MN (US); Gary Berosik, South St. Paul, MN (US); and Thomas Zielund, Shakopee, MN (US)
Assigned to Thomson Reuters Global Resources, Baar (CH)
Filed by Thomson Reuters Global Resources, Baar (CH)
Filed on Jul. 21, 2014, as Appl. No. 14/337,105.
Application 14/337,105 is a continuation of application No. 12/553,752, filed on Sep. 3, 2009, granted, now 8,788,523.
Application 12/553,752 is a continuation of application No. 12/354,617, filed on Jan. 15, 2009, abandoned.
Claims priority of provisional application 61/011,147, filed on Jan. 15, 2008.
Prior Publication US 2015/0112877 A1, Apr. 23, 2015
This patent is subject to a terminal disclaimer.
Int. Cl. G06F 17/30 (2006.01); G06Q 10/00 (2012.01); G06F 17/27 (2006.01); G06Q 10/06 (2012.01); G06Q 10/10 (2012.01); G06Q 50/18 (2012.01); G06F 15/16 (2006.01); G06F 17/00 (2006.01); G06Q 50/00 (2012.01); G06F 17/24 (2006.01)
CPC G06F 17/2775 (2013.01) [G06F 15/16 (2013.01); G06F 17/00 (2013.01); G06F 17/243 (2013.01); G06F 17/2705 (2013.01); G06F 17/30 (2013.01); G06F 17/30011 (2013.01); G06F 17/30194 (2013.01); G06F 17/30227 (2013.01); G06F 17/30244 (2013.01); G06Q 10/06 (2013.01); G06Q 10/10 (2013.01); G06Q 50/00 (2013.01); G06Q 50/18 (2013.01)] 19 Claims
OG exemplary drawing
 
1. A method comprising:
processing a physical document by capturing an image representation of the physical document and generating a set of image data associated with the physical document;
transforming the set of image data into a set of electronic text representing text appearing on the physical document;
receiving the set of electronic text;
parsing by a phrase discovery engine the electronic text into a set of tokens and identifying and extracting a set of legal clauses, wherein each legal clause in the set of legal clauses comprises two or more semi-contiguous tokens;
comparing the set of legal clauses to legal clauses previously derived from a corpus of other electronic documents, wherein comparing the set of legal clauses to previously derived legal clauses from a corpus of other electronic documents comprises providing an index of clauses, wherein each clause in the index of clauses is associated with a legal classification from a set of legal classifications;
identifying one or more legal clauses from the set of legal clauses based on the comparison; and
assigning a legal classification to one or more of the identified legal clauses in the identified set of legal clauses based on the set of legal classifications.