System for searching, recommending, and exploring documents through conceptual associations

Patent metadata

Field	Value
Publication number	US-2016012126-A1
Application number	US-201514657391-A
Country	US
Kind code	A1
Filing date	Mar 13, 2015
Priority date	Jul 14, 2014
Publication date	Jan 14, 2016
Grant date	—

Abstract

According to an aspect, searching, recommending, and exploring documents through conceptual associations includes a method for receiving a plurality of documents and extracting concepts from each of the documents. A degree of relation between each of the documents and concepts in a knowledge base is calculated. The method also includes, in response to receiving a query, determining one or more concepts from the query. For each of the concepts, a list of documents having a highest degree of relation to the concept is retrieved. The method also includes outputting a list that is responsive to the one or more retrieved lists.

First claim

What is claimed is:

1 . A method comprising:
receiving a plurality of documents; extracting concepts from each of the documents; computing a degree of relation between each of the documents and concepts in a knowledge base; and in response to receiving a query:
determining one or more concepts from the query;
for each of the determined concepts, retrieving a list of documents having a highest degree of relation to the concept; and
outputting a list responsive to the one or more retrieved lists.

2 . The method of claim 1 , wherein the outputting comprises:
upon identification of one concept resulting from the determining, outputting the corresponding list; and upon identification of two or more concepts resulting from the determining, merging corresponding lists into a single list, and outputting the merged list.

3 . The method of claim 1 , further comprising:
determining that one of the documents has been updated; and performing the extracting and computing for the document based on the determining.

4 . The method of claim 1 , wherein a subject and content of each of the documents are associated with a corresponding person, and the extracted concepts include attributes of the person, wherein one or more documents in the list indicate at least one of:
a person estimated to be interested in the query based on a degree of relation of the one or more documents to the concept of the query; and a person estimated to have expertise in the concepts described in the query.

5 . The method of claim 1 , further comprising storing results of the computing a degree of relation in a conceptual inverted index.

6 . The method of claim 1 , wherein the outputting the list includes displaying, via a user interface, highlighted extracted concepts in the document that are determined to have the highest degree of relation to the concept.

7 . The method of claim 1 , wherein the knowledge base is a concept graph.

8 . The method of claim 7 , wherein the computing uses the extracted concepts and the concept graph.

9 . The method of claim 1 , wherein the extracting concepts from each of the documents includes computing a probability that text in the document is generated with two or more language models, and performing a differential analysis on the probability.

10 . The method of claim 1 , wherein the query is input as text, and concepts are extracted from the text to create a conceptual query.

11 . The method of claim 1 , wherein the computing is responsive to paths connecting the extracted concepts to the concepts in the knowledge base.

Assignees

• IBM

Inventors

• Franceschini Michele M
• Lastras-Montano Luis A
• Richards John T
• Soares Livio B
• Wegman Mark N

Classifications

• G06F16/3344

  using natural language analysis · CPC title

• G06F16/3346

  using probabilistic model · CPC title

• G06F16/325 Primary

  Hash tables · CPC title

• G06F16/951

  Indexing; Web crawling techniques · CPC title

• G06F17/30864

  Physics · mapped topic