System and method for automatic semantic labeling of natural language texts

• US 8,583,422 B2
• Filed: 03/12/2010
• Issued: 11/12/2013
• Est. Priority Date: 03/13/2009
• Status: Active Grant

• Current Assignee: Allium US Holding LLC
• Original Assignee: Invention Machine Corporation (S&P Global Inc.)
• Inventors: Todhunter, James, Sovpel, Igor, Pastanohau, Dzianis
• Primary Examiner(s): He, Jialong
• Application Number:
  US12/723,472
• Publication Number:
  US 20100235165A1
• Time in Patent Office: 1,341 Days
• Field of Search: 704/1, 704/9, 707/1, 707/706, 707/759, 715/205, 715/235, 715/853, 715/854
• US Class Current: 704/9
• CPC Class Codes: G06F 16/3329 Natural language query formulation or dialogue systems
• CPC Class Codes: G06F 40/35 Discourse or dialogue representation

Abstract

Systems and methods for automatic semantic labeling of natural language documents provided in electronic or digital form include a semantic processor that performs a basic linguistic analysis of text, including recognizing in the text semantic relationships of the type objects and/or classes of objects, facts and cause-effect relationships; matching linguistically analyzed text against target semantic relationship patterns, created by generalization of particular cases of target semantic relationships; and generating semantic relationship labels based on linguistically analyzed text and a result of the matching.

Claims

1. A processor-implemented method for automatic labeling of natural language text, the method comprising:

receiving text from at least one natural language document in electronic form;

performing, using a processor, a basic linguistic analysis of the text that includes recognizing cause-effect relationships in the text and generating cause-effect labels for words or phrases in the text that form part of the cause-effect relationships;

matching the linguistically analyzed text and the generated cause-effect labels against stored target semantic relationship patterns, wherein the stored target semantic relationship patterns generically describe semantic relationships between words or phrases, the stored target semantic relationships being derived in part from cause-effect relationships between words or phrases;

producing additional semantic relationship labels for the linguistically analyzed text based on the matching of the linguistically analyzed text and the generated cause-effect labels against the stored target semantic relationship patterns, wherein the additional semantic relationship labels are tagged to words or phrases from sentences within the linguistically analyzed text in order to identify semantic relationships between those words or phrases by identifying those words or phrases as components of semantic relationships of the stored target semantic relationship patterns; and

storing the linguistically analyzed text and the additional semantic relationship labels in a non-transitory storage medium.

2. The method according to claim 1, further comprising:

applying parts-of-speech tags to at least portions of the text to generate tagged portions of the text;

parsing the tagged portions of the text to generate parsed and tagged portions of the text; and

wherein performing, using the processor, the basic linguistic analysis of the text includes semantically analyzing the parsed and tagged portions of the text to generate semantically analyzed, parsed and tagged portions of the text.

3. The method according to claim 2, wherein applying parts-of-speech tags is performed on preformatted portions of the text, whereby the preformatted portions of the text comprise the text with non-natural language symbols removed.

4. The method according to claim 2, wherein semantically analyzing the parsed and tagged portions of the text comprises:

recognizing one or more facts in the form of at least one expanded Subject-Action-Object (eSAO) set in the text, wherein each eSAO set has at least one eSAO component; and

recognizing in the text a set of rules that reflect regularities of at least one of an external domain and a knowledge domain in the form of cause-effect relationships in at least one eSAO set, wherein at least one cause-effect relationship of the cause-effect relationships comprises a cause eSAO and an effect eSAO.

5. The method according to claim 4, wherein the at least one eSAO component comprises text related to one or more elements selected from the group consisting of subjects, objects, actions, adjectives, prepositions, indirect objects, and adverbs.

6. The method according to claim 4, wherein recognizing one or more expanded Subject-Action-Object (eSAO) sets in the text comprises recognizing one or more subjects, objects, actions, adjectives, prepositions, indirect objects, and adverbs in at least one sentence of the text.

7. The method according to claim 4, wherein recognizing one or more expanded Subject-Action-Object (eSAO) sets and cause-effect relationships in the text comprises accessing a linguistic knowledge base having a database of patterns defining eSAO and cause-effect components.

8. The method according to claim 4, wherein the cause eSAO comprises at least one eSAO component of the at least one eSAO set and the effect eSAO comprises at least one other eSAO component of the at least one eSAO set.

9. The method according to claim 8, wherein the at least one cause-effect relationship comprises a sequential operator relating the at least one eSAO component of the cause eSAO to the at least one other eSAO component of the effect eSAO with lexical, grammatical, and/or semantic language means.

10. The method according to claim 1, wherein matching the linguistically analyzed text and the generated cause-effect labels against stored target semantic relationship patterns that are derived in part from cause-effect relationships further comprises:

accessing a semantic relationship pattern database that is a part of a linguistic knowledge database, wherein the semantic relationship pattern database is generated by;

performing a basic linguistic analysis of a corpus of text documents;

recognizing in the linguistically analyzed corpus particular cases of target semantic relationships;

generalizing the particular cases of target semantic relationships into semantic relationship linguistic patterns using lexical language units and their semantic classes, part-of-speech and syntactic tags, eSAO and cause-effect labels from the recognized particular cases of target semantic relationships; and

storing the semantic relationship linguistic patterns.

11. The method according to claim 10, wherein generalizing the particular cases of target semantic relationships into semantic relationship linguistic patterns uses an eSAO format as a context, and comprises:

generalizing constituent components of the particular cases of target semantic relationships by searching in the linguistically analyzed corpus of text documents using lexical, grammatical, syntactic, eSAO and cause-effect labels obtained for the components from the basic linguistic analysis.

12. The method according to claim 1, wherein matching the linguistically analyzed text and the generated cause-effect labels against the stored target semantic relationship patterns comprises matching the linguistically analyzed text against words, part-of-speech tags, syntactic tags, eSAO, and cause-effect sets in the stored target semantic relationship patterns.

13. A semantic processor for automatically semantic labeling of natural language text in electronic or digital form, the semantic processor comprising:

a preformatter that preformats received electronic text;

a linguistic analyzer that performs basic linguistic analysis of the preformatted text using a central processing unit (CPU), wherein the basic linguistic analysis includes;

recognizing objects or classes of objects within the text;

recognizing facts over the objects or classes of objects;

recognizing cause-effect relationships over the facts; and

creating tags that associate the text with corresponding cause-effect relationships; and

a labeler that matches the linguistically analyzed text and the created tags against stored target semantic relationship patterns to produce semantically labeled text, wherein the stored target semantic relationship patterns generically describe semantic relationships between words or phrases, the stored target semantic relationships being derived in part from cause-effect relationships between words or phrases, wherein the semantically labeled text includes semantic relationship labels that indicate semantic relationships between words or phrases from sentences within the text, wherein the semantic relationship labels indicate which words or phrases in the linguistically analyzed text are components of semantic relationships of the stored target semantic relationship patterns.

14. The semantic processor according to claim 13, wherein the linguistic analyzer comprises a semantic analyzer that produces semantically analyzed text, the semantic analyzer comprising:

an expanded Subject-Action-Object (eSAO) recognizer that recognizes eSAOs sets in the text; and

a cause-effect (C-E) recognizer that recognizes an eSAO cause-effect relationship of the cause-effect relationships,

wherein eSAO and C-E recognition is based on linguistic patterns stored in a linguistic knowledge base.

15. The semantic processor according to claim 14, wherein the eSAO cause-effect relationship comprises a cause eSAO, an effect eSAO, and at least one sequential operator relating the cause eSAO to the effect eSAO.

16. The semantic processor according to claim 15, wherein each eSAO set comprises eSAO components and the cause eSAO comprises at least one eSAO component of the eSAO components and the effect eSAO comprises at least one eSAO component of the eSAO components that is different from the at least one eSAO component of the cause eSAO.

17. The semantic processor according to claim 14, wherein the eSAO components comprise text related to one or more elements selected from the group consisting of subjects, objects, actions, adjectives, prepositions, indirect objects and adverbs.

18. The semantic processor according to claim 14, wherein the linguistic analyzer further comprises:

a part-of-speech (POS) tagger that receives the preformatted text and produces POS tagged text; and

a parser that receives the POS tagged text, produces parsed text, and provides the parsed text to the semantic analyzer,

wherein the parts-of-speech tagger and the parser operate with data stored in the linguistic knowledge base.

19. The semantic processor according to claim 13, wherein the preformatter performs at least one of a removal of any symbols in a digital or electronic presentation of the text that do not form part of natural language text, a detection and correction of any mismatches or mistakes in the text, and partitioning the text into structures of sentences and words.

20. The semantic processor according to claim 13, wherein the stored target semantic relationship patterns are created by a pattern generator comprising:

a corpus linguistic analyzer that performs basic linguistic analysis of a corpus of text documents;

a labeled text corpus generator that generates a labeled text corpus having part-of-speech tags, syntactic tags, eSAO labels, and cause-effect labels;

a relation generator that recognizes in the labeled text corpus particular cases of target semantic relationships;

a pattern generator that generalizes the particular cases of semantic relationships by using their labels to generate more general semantic relationship linguistic patterns, wherein the labels include lexical language units, their semantic classes, part-of-speech and syntactic tags, and eSAO and cause-effect labels; and

a pattern tester for testing the general semantic relationship linguistic patterns by the pattern generator.

21. The semantic processor according to claim 20, wherein when generalizing the particular cases of target semantic relationships into semantic relationship linguistic patterns, the pattern generator uses an eSAO format as a context to:

generalize constituent components as a result of searching in the linguistically analyzed corpus of text documents using of part-of-speech, syntactic, and eSAO and cause-effect labels obtained for the components at a level of the basic linguistic analysis.

22. The semantic processor according to claim 13, wherein the labeler matches the linguistically analyzed text against target semantic relationship patterns by matching words, part-of-speech tags, syntactic tags, eSAO and cause-effect sets.

23. A computer-implemented method for recognizing new types of linguistic patterns for use in a pattern database, the method comprising,

performing, using a processor, a first level linguistic analysis of a natural language document to recognize and tag words indicative of objects and actions in the natural language document;

performing a second level linguistic analysis of the natural language document by comparing the tags created during the first level linguistic analysis to linguistic patterns stored in a linguistic knowledge database and by creating semantic relationship labels indicative of semantic relationship linguistic patterns in the natural language document, the semantic relationship linguistic patterns including cause-effect relationships patterns; and

performing a third level linguistic analysis of the natural language document by recognizing a new type of semantic relationship linguistic pattern based on the semantic relationship labels created during the second level linguistic analysis and on target semantic relationship patterns that generically describe semantic relationships between words or phrases, the target semantic relationships being derived in part from cause-effect relationships between words or phrases, the new type of semantic relationship linguistic pattern identifying a new semantic relationship between words or phrases in the natural language document.

24. The computer-implemented method of claim 23, wherein:

performing the first level linguistic analysis includes recognizing and tagging components of at least one eSAO set;

performing the second level linguistic analysis includes creating labels indicative of cause-effect patterns in the natural language document, the labels identifying words or phrases in the natural language document as components of the cause-effect patterns; and

performing the third level linguistic analysis includes recognizing the new type of semantic relationship linguistic pattern using the labels indicative of cause-effect patterns created during the second level linguistic analysis, wherein the new type of semantic relationship linguistic pattern identifies a new semantic relationship between the words or phrases that are the components of the cause-effect patterns.