Word Deletion for Searches

Abstract

A searcher can be configured to improve search results through the use of intelligent word deletion. A search auto categorizer (SAC) operates on the original query and returns a list of leaf categories and a distribution of probabilities among the leaf categories. The original query is parsed into word pairs and each word pair is run through the search engine. The search results for each word pair are weighted by the leaf category probabilities. A word pair is selected from the results and one of the two words is deleted from the original query. The searcher can perform exhaustive deletion where multiple truncated queries are generated from the original query and the results list from one truncated query is returned as the results list. The searcher can build up a truncated query from the original query by iteratively appending a word selected from the original query to the truncated query.

1 . In a computerized search system which queries are submitted by users who receive, in response, a list of documents selected from a corpus of documents wherein the list comprises documents deemed responsive to a user's query, a method of preprocessing the query comprising:
 obtaining a base query from a user input, wherein the base query comprises a plurality of words;   generating a plurality of truncated queries, each truncated query comprising a portion of the base query from which at least one word is removed;   running each of the plurality of truncated queries against the corpus of documents to obtain a count value of a number of search results in the document corpus deemed responsive to the truncated query;   comparing the count values obtained for each of the truncated queries;   selecting one of the truncated queries based on the comparison;   modifying the base query with information from the selected truncated query; and   running the modified base query against the corpus of documents to obtain a list of documents deemed responsive to the user's query.   
     
     
         2 . The method of  claim 1 , wherein generating the plurality of truncated queries comprises generating all possible truncated queries having one word removed from the base query. 
     
     
         3 . The method of  claim 1 , wherein selecting the truncated query comprises selecting the truncated query having a lowest non-zero count value. 
     
     
         4 . The method of  claim 1 , wherein selecting the truncated query comprises selecting the truncated query having a highest count value. 
     
     
         5 . The method of  claim 1 , further comprising:
 determining whether the count value for each of the plurality of truncated queries is zero;   generating a new truncated query by selecting a single word from the plurality of words;   generating an updated truncated query by successively appending the new truncated query with a distinct word selected from the plurality of words so long as a count value associated with the updated truncated query is non-zero; and   running the updated truncated query against the corpus of documents to obtain a buildup results list of one or more documents in the document corpus deemed responsive to the updated truncated query.   
     
     
         6 . In a computerized search system which queries are submitted by users who receive, in response, a list of documents selected from a corpus of documents wherein the list comprises documents deemed responsive to a user's query, a method of preprocessing the query comprising:
 obtaining a base query from a user input, wherein the base query comprises a plurality of words;   generating a truncated query by selecting a single word from the plurality of words of the base query;   determining candidate words from the plurality of words, each candidate word having a non-zero count value of a number of results when appended to the truncated query;   selecting a distinct candidate word based on its corresponding count value compared to the count values of the other candidate words and an optimization policy of the computerized search system;   generating an updated truncated query by appending the truncated query with the distinct candidate word; and   running the updated truncated query against the corpus of documents to obtain a results list of one or more documents in the document corpus deemed responsive to the updated truncated query.   
     
     
         7 . The method of  claim 6 , further comprising:
 repeating the act of generating a further updated query by appending the updated truncated query with an additional distinct candidate word to further optimize the count value until the results list contains no documents; and   returning a result list corresponding to a latest non-zero count value.   
     
     
         8 . The method of  claim 6 , wherein selecting the single word from the plurality of words comprises selecting a word resulting in a minimum non-zero count value for the truncated query. 
     
     
         9 . The method of  claim 6 , wherein selecting the single word from the plurality of words comprises selecting a word resulting in a maximum count value for the truncated query. 
     
     
         10 . The method of  claim 6 , wherein the distinct candidate word is selected to minimize the count value for the updated truncated query. 
     
     
         11 . The method of  claim 6 , wherein the distinct candidate word is selected to maximize the count value for the updated truncated query. 
     
     
         12 . In a computerized search system which queries are submitted by users who receive, in response, a list of documents selected from a corpus of documents wherein the list comprises documents deemed responsive to a user's query, an apparatus for preprocessing the query comprising:
 a search engine configured to receive a base query from a user input, wherein the base query comprises a plurality of words;   a Search Logic Unit (SLU) configured to generate a plurality of truncated queries, each truncated query comprising a portion of the base query from which at least one word is removed, the SLU submitting each of the plurality of truncated queries as an input to the search engine to obtain a count value of a number of search results in the document corpus deemed responsive to each of the truncated queries; and   wherein the SLU compares the count values obtained for each of the truncated queries, selects one of the truncated queries based on the comparison and modifies the base query with information from the selected truncated query, and wherein the search engine runs the modified base query against the corpus of documents to obtain a list of documents deemed responsive to the user's query.   
     
     
         13 . The apparatus of  claim 12 , wherein the SLU selects the truncated query having a lowest non-zero count value. 
     
     
         14 . The apparatus of  claim 12 , wherein the SLU selects the truncated query having a greatest count value. 
     
     
         15 . The apparatus of  claim 12 , wherein the search engine returns a null results list when the count value corresponding to all of the truncated queries is zero. 
     
     
         16 . In a computerized search system which queries are submitted by users who receive, in response, a list of documents selected from a corpus of documents wherein the list comprises documents deemed responsive to a user's query, an apparatus for preprocessing the query comprising:
 a search engine configured to receive a base query from a user input, wherein the base query comprises a plurality of words;   a Search Logic Unit (SLU) configured to generate a truncated query by selecting a single word from the plurality of words, determine candidate words from the plurality of words, each candidate word having a non-zero count value of a number of results when appended to the truncated query, select a distinct candidate word based on its corresponding count value compared to the count values of the other candidate words and an optimization policy of the computerized search system, generate an updated truncated query by appending the truncated query with the distinct candidate word; and   wherein the SLU submits the updated truncated query to the search engine to run against the corpus of documents to obtain a results list of one or more documents in the document corpus deemed responsive to the updated truncated query.   
     
     
         17 . The apparatus of  claim 16 , wherein the SLU is configured to select the distinct candidate word for appending to the truncated query that results in a minimum count value for the updated truncated query. 
     
     
         18 . The apparatus of  claim 16 , wherein the SLU is configured to select the distinct candidate word for appending to the truncated query that results in a maximum count value for the updated truncated query.