Advanced text tagging using key phrase extraction and key phrase generation

Abstract

The systems and methods described herein describe a comprehensive knowledge and/or skills management technology tool to address these and other issues with an advanced text tagging algorithm to extract the relevant topics from a text segment. The tagging algorithm includes a key phrase extraction technique and a key phrase generation technique. The key phrase extraction includes identifying phrases from the original text that represents its most relevant information. The key phrase generation technique includes generating additional phrases that do not necessarily appear in the text, but which describe its subject.

1 .- 14 . (canceled) 
     
     
         15 . A system configured to tag text in a document for use in a knowledge and/or skills management technology platform, the system comprising:
 a computer system comprising one or more physical processors programmed with computer program instructions that, when executed by the one or more physical processors, program the computer system to:   execute a tagging algorithm to extract relevant topics from a text segment, the tagging algorithm comprising a key phrase extraction component and a key phrase generation component, wherein executing the tagging algorithm comprises:   extracting, by the key phrase extraction component, key phrase candidates in the text segment that represent its most relevant information;   generating, by the key phrase generation component one or more key phrase candidates that do not appear in the text segment; and   applying a machine learning algorithm to create and store a knowledge and/or skills map,   that links a user to a topic indicated by one or more tag suggestions and that indicates a user score for the user in relation to the topic.   
     
     
         16 . The system of  claim 15  wherein the key phrase extraction algorithm is based on part-of-speech tagging, where a part-of-speech tag identifies the grammatical role of words in a sentence, the algorithm being configured to obtain tokens corresponding to a sentence and obtain a part-of-speech tag for the tokens. 
     
     
         17 . The system of  claim 15  wherein the key phrase extraction algorithm identifies key phrase candidates, filters the key phrase candidates, generates a score for the candidates and selects the most relevant candidates. 
     
     
         18 . The system of  claim 15  wherein the key phrase extraction algorithm is configured to score the candidates based on a set of stored criteria, including:
 (i) the frequency of occurrences of the phrase in the text; 
 (ii) the position in the text; 
 (iii) a determination of the appearance of the phrase in an online encyclopedic database; 
 (iv) the commonality of the phrase in a network associated with the knowledge and/or skills management technology platform; and/or 
 (v) the lack of commonality of the phrase in the language in general; and 
 generate a tag suggestion including the top-scored phrases and the generated key phrase. 
 
     
     
         19 . The system of  claim 15 , wherein the key phrase generation algorithm is configured to: derive suggested hypernyms from a database of potential hyponym-hypernym pairs; and for a given set of key phrases, the algorithm identifies a hypernym candidate if there are at least two associated hyponyms among the set of key phrases; and selects a set of key phrases based on a determination of relevance. 
     
     
         20 . The system of  claim 19  wherein the algorithm is configured to determine relevance based on at least the following factors: the total input key phrase coverage; the total occurrences of the hyponyms in the language; and the total page views of their corresponding online encyclopedia page; and sort the candidates sequentially and add a hypernym in the final suggestion if at least one of the hyponyms covered by the candidate was not covered by any of the previously suggested hypernyms. 
     
     
         21 . The system of  claim 15  wherein the key phrase generation algorithm is configured to suggest hypernyms, including: tag suggestion hypernyms based on key phrases suggested from the key phrase extraction; and/or user-context hypernyms based on a user's top-linked tags. 
     
     
         22 . A method for tagging text in a document for use in a knowledge and/or skills management technology platform, the method comprising:
 programming a computer system comprising one or more physical processors with computer program instructions that, when executed by the one or more physical processors, program the computer system to perform the steps of:   executing a tagging algorithm to extract relevant topics from a text segment, the tagging algorithm comprising a key phrase extraction component and a key phrase generation component, wherein executing the tagging algorithm comprises:   extracting, by the key phrase extraction component, key phrase candidates from in the text segment that represent its most relevant information;   generating, by the key phrase generation component one or more key phrase candidates that do not appear in the text segment; and   applying a machine learning algorithm to create and store a knowledge and/or skills map,   that links a user to a topic indicated by one or more tag suggestions and that indicates a user score for the user in relation to the topic.   
     
     
         23 . The method of  claim 22  wherein the key phrase extraction algorithm is based on part-of-speech tagging, where a part-of-speech tag identifies the grammatical role of words in a sentence, the algorithm being configured to obtain tokens corresponding to a sentence and obtain a part-of-speech tag for the tokens. 
     
     
         24 . The method of  claim 22  wherein the key phrase extraction algorithm identifies key phrase candidates, filters the key phrase candidates, generates a score for the candidates and select the most relevant candidates. 
     
     
         25 . The method of  claim 22  wherein the key phrase extraction algorithm is configured to score the candidates based on a set of stored criteria, including:
 (i) the frequency of occurrences of the phrase in the text; 
 (ii) the position in the text; 
 (iii) a determination of the appearance of the phrase in an online encyclopedic database; 
 (iv) the commonality of the phrase in a network associated with the knowledge and/or skills management technology platform; and/or 
 (v) the lack of commonality of the phrase in the language in general; and generate a tag suggestion including the top-scored phrases and the generated key phrase. 
 
     
     
         26 . The method of  claim 22 , wherein the key phrase generation algorithm is configured to: derive suggested hypernyms from a database of potential hyponym-hypernym pairs; and for a given a set of key phrases, the algorithm identifies a hypernym candidate if there are at least two associated hyponyms among the set of key phrases; and selects a set of key phrases based on a determination of relevance. 
     
     
         27 . The method of  claim 26  wherein the algorithm is configured to determine relevance based on at least the following factors: the total input key phrase coverage; the total occurrences of the hyponyms in the language; and the total page views of their corresponding online encyclopedia page; and sort the candidates sequentially and add a hypernym in the final suggestion if at least one of the hyponyms covered by the candidate was not covered by any of the previously suggested hypernyms. 
     
     
         28 . The method of  claim 22  wherein the key phrase generation algorithm is configured to suggest hypernyms, including: tag suggestion hypernyms based on key phrases suggested from the key phrase extraction; and/or user-context hypernyms based on a user's top-linked tags.