发明名称 SYSTEMS AND METHODS FOR THE CREATION, UPDATE AND USE OF MODELS IN FINDING AND ANALYZING CONTENT
摘要 Embodiments of a search system that provides knowledge based searching of content based on a knowledge model created from the content being searched are disclosed. Embodiments of such search systems may build a model of entitles and relationships representing the collective knowledge contained in a set of content analyzed. When a search is performed on content the model may be leveraged to improve the accuracy, relevance and recall of the search.
申请公布号 US2017011092(A1) 申请公布日期 2017.01.12
申请号 US201615205758 申请日期 2016.07.08
申请人 Trendkite Inc. 发明人 Huddleston Erik Lee;Perdue David Francis;Allison Matthew John;De Oliveira John Joseph
分类号 G06F17/30;G06N5/02 主分类号 G06F17/30
代理机构 代理人
主权项 1. A searching system providing knowledge based searching of content based on a knowledge model created from the content being searched, comprising: a first data store, including: content comprised of a set of articles, and an index having an entry for each of the set of articles; a second data store, including: an ontology modeling entities and relationships of the content, where each relationship has an associated strength value determined based on the content; an entity extractor for: receiving an article from a content source over a network,determining a set of entities from the article, anddetermining proximity data associated with the set of entities from the article, wherein the proximity data includes relative positional information for the entities; an entity evaluator for: receiving the set of entities of the article and the proximity data,determining an entity score for each of the set of entities,saving each of the set of entities and the associated entity score in an entry in the index associated with the article; a relationship evaluator for: receiving the set of entities of the article and the proximity data,determining a set of relationships between each of the set of entities of the article,determining an article strength for each of the set of relationships based on the proximity data associated with the set of entities,updating the ontology, wherein updating the ontology comprises updating the strength of each relationship of the set of relationships in the ontology based on the article strength for that relationship; and a search interface for: receiving one or more search entities through the search interface,determining an initial set of articles based on the index, wherein for each of the initial set of articles the entry in the index associated with that article includes at least one of the search entities,determining a relevancy score for each of the initial set of articles using the entity score for each of the one or more search entities in the entry in the index for that article and the relationship strength for each relationship between any of the one or more search entities from the ontology, ranking the initial set of articles by the relevancy score, and returning the ranked set of articles through the search interface.
地址 Austin TX US