Semantic search tool for document tagging, indexing and search

Patent metadata

Field	Value
Publication number	US-9684683-B2
Application number	US-201314051984-A
Country	US
Kind code	B2
Filing date	Oct 11, 2013
Priority date	Feb 9, 2010
Publication date	Jun 20, 2017
Grant date	Jun 20, 2017

Abstract

A semantic tagging engine automatically generates semantic tags for the given documents and enables semantic search, based on meanings of search terms and content tags. A Semantic Knowledge Management Tool (SKMT) forms a semantic search and knowledge management platform to search, analyze and manage enterprise content. SKMT scans different content sources and generates indexes of semantic keywords. Its interface allows users to manage various data sources, search, explore and visualize search results at semantic level. SKMT provides high precision of semantic search and semantic data visualization.

First claim

The invention claimed is:

1. A method for performing a semantic search to retrieve documents from a document repository, comprising:
a processor accepting through a graphical interface a search phrase provided by a user;
the processor analyzing the search phrase and suggesting a plurality of semantic search phrases generated from a stored vocabulary of an ontology to suggest a context;
the processor accepting a first semantic search phrase selected from the plurality of semantic search phrases;
the processor indexing and semantically tagging a plurality of documents in a storage medium, wherein at least one document was unstructured prior to the indexing and semantically tagging, wherein indexing and semantically tagging the plurality of documents comprises the processor weighing a plurality of semantic concepts generated from a plurality of extracted noun phrases, wherein the plurality of semantic concepts is weighed by term significance (TS) scores, wherein the TS scores are calculated with both semantic and statistical information, and wherein the semantic information includes lexical chaining Word Sense Disambiguation (WSD) scores and Depth and Information Content (IC) values, and the statistical information includes Term Frequency (TF) and Inverse Google Popularity (IGP);
the processor extracting a plurality of semantic tags from the indexed and semantically tagged documents; and
the processor defining a plurality of groups based on the indexed and semantically tagged documents, wherein the documents in the plurality of documents are indexed with unique identifications of semantic tags, and wherein the processor searches the indexed documents based on a unique identification of the first semantic search phrase.

2. The method of claim 1 , wherein the at least one unstructured document is stored in a local file and is semantically indexed before the search phrase is accepted by the processor.

3. The method of claim 1 , wherein the processor retrieves the at least one unstructured document via an Internet and semantically tags and indexes the retrieved document after the search phrase is accepted by the processor.

4. The method of claim 1 , wherein the semantically tagging enables a structured query search of the document that was unstructured.

5. The method of claim 1 , wherein at least one group of the plurality of groups is determined by a keyword that occurs in a selected document in the storage medium.

6. The method of claim 1 , wherein at least one group of the plurality of groups is determined by a semantic tag that occurs in a selected document in the storage medium.

7. The method of claim 6 , wherein the at least one group is represented in a semantic tag cloud.

8. The method of claim 1 , further comprising: selecting a group in the plurality of groups; and updating the semantic search in accordance with the selected group.

9. A system to perform a semantic search to retrieve documents from a document repository, comprising:
a memory configured to store and retrieve data, including instructions; a processor configured to execute instructions retrieved from the memory to perform the steps:
accepting a search phrase; analyzing the search phrase and suggesting a plurality of semantic search phrases generated from a stored vocabulary of an ontology to suggest a context;
accepting a first semantic search phrase selected from the plurality of semantic search phrases;
storing in a storage medium at least one unstructured document;
indexing and semantically tagging a plurality of documents including the at least one unstructured document in the storage medium, wherein indexing and semantically tagging the plurality of documents comprises weighing a plurality of semantic concepts generated from a plurality of extracted noun phrases, wherein the plurality of semantic concepts is weighed by term significance (TS) scores, wherein the TS scores are calculated with both semantic and statistical information, and wherein the semantic information includes lexical chaining Word Sense Disambiguation (WSD) scores and Depth and Information Content (IC) values, and the statistical information includes Term Frequency (TF) and Inverse Google Popularity (IGP);
extracting a plurality of semantic tags from the indexed and semantically tagged documents; and
defining one or more groups based on the indexed and semantically tagged documents, wherein the documents in the plurality of documents are indexed with unique identifications of semantic tags, and wherein the processor searches the indexed documents based on a unique identification of the first semantic search phrase.

10. The system of claim 9 , wherein documents are stored in a local file and the documents are semantically indexed before the search phrase is accepted by the processor.

11. The system of claim 9 , wherein the processor retrieves the at least one unstructured document via an Internet based on the search phrase and semantically tags and indexes the at least one unstructured document to make it a structured document.

12. The system of claim 9 , wherein the semantically tagging enables a structured query search of the at least one document that was previously unstructured.

13. The system of claim 9 , wherein at least one group of the plurality of groups is determined by a keyword that occurs in a selected document in the storage medium.

14. The system of claim 9 , wherein at least one group of the plurality of groups is determined by a semantic tag that occurs in a selected document in the storage medium.

15. The system of claim 14 , wherein the at least one group is represented in a semantic tag cloud.

16. The system of claim 9 , further comprising the steps performed by the processor: selecting a group in the plurality of groups; and updating the semantic search in accordance with the selected group.

Assignees

• Dang Jlangbo
• Kalender Murat
• Toklu Candemir
• Hampel Kenneth
• Siemens Ag

Inventors

• Dang Jlangbo
• Kalender Murat
• Toklu Candemir
• Hampel Kenneth

Classifications

• G06F16/31

  Indexing; Data structures therefor; Storage structures · CPC title

• G06F16/3322

  using system suggestions (G06F16/3325 takes precedence) · CPC title

• G06F16/2228 Primary

  Indexing structures · CPC title

• G06F16/345

  Summarisation for human users · CPC title

• G06F40/30 Primary

  Semantic analysis · CPC title