Search clustering

Abstract

In one example embodiment, a method is illustrated as including retrieving item data. At least one base cluster having at least one document with common item data stored in a suffix ordering is constructed. The at least one base cluster is compacted to create a compacted cluster representation having a reduced duplicate suffix ordering amongst the clusters.

1 . (canceled) 
     
     
         2 . A machine-readable storage medium that does not comprise transitory signals, the machine-readable storage medium storing instructions which, when executed by at least one processor of a machine, cause the machine to perform operations comprising:
 retrieving, from a database, item data;   constructing at least one base cluster having at least one document with common item data stored in a suffix ordering; and   compacting the at least one base cluster to create a compacted cluster representation having a reduced duplicate suffix ordering amongst clusters.   
     
     
         3 . The machine-readable storage medium of  claim 2 , wherein the operations further comprise merging the compacted cluster representation to generate a merged cluster based on a first overlap value applied to the at least one document with the common item data. 
     
     
         4 . The machine-readable storage medium of  claim 2 , wherein the operations further comprise identifying noise data using a demand factor that is based on relationships of items and categories to query terms of search queries, wherein the retrieving the item data comprises filtering the noise data from the item data. 
     
     
         5 . The machine-readable storage medium of  claim 2 , wherein the operations further comprise filtering noise data from item data based on a frequency with which a word is used in a search query as compared to a frequency another word is used in the search query. 
     
     
         6 . The machine-readable storage medium of  claim 2 , wherein the item data is part of a plurality of merged clusters organized into a hierarchy of merged clusters. 
     
     
         7 . The machine-readable storage medium of  claim 2 , wherein the operations further comprise merging the compact cluster representation based on at least one selection from the group consisting of a relevance-weight factor, a seller factor, a price factor, a category factor, and an image factor. 
     
     
         8 . The machine-readable storage medium of  claim 2 , wherein the operations further comprise calculating a relevance score using a demand factor. 
     
     
         9 . The machine-readable storage medium of  claim 8 , wherein the relevance score is calculated, in part, based on a comparison of a similarity of a demand category histogram and a supply category histogram. 
     
     
         10 . The machine-readable storage medium of  claim 8 , wherein the operations further comprise using the relevance score to prune irrelevant clusters. 
     
     
         11 . A method comprising:
 retrieving, from a database, item data;   constructing at least one base cluster having at least one document with common item data stored in a suffix ordering; and   compacting, using a hardware processor, the at least one base cluster to create a compacted cluster representation having a reduced duplicate suffix ordering amongst clusters.   
     
     
         12 . The method of  claim 11 , further comprising merging the compacted cluster representation to generate a merged cluster based on a first overlap value applied to the at least one document with the common item data. 
     
     
         13 . The method of  claim 11 , further comprising identifying noise data using a demand factor that is based on relationships of items and categories to query terms of search queries, wherein the retrieving the item data comprises filtering the noise data from the item data. 
     
     
         14 . The method of  claim 11 , further comprising filtering noise data from item data based on a frequency with which a word is used in a search query as compared to a frequency another word is used in the search query. 
     
     
         15 . The method of  claim 11 , wherein the item data is part of a plurality of merged clusters organized into a hierarchy of merged clusters. 
     
     
         16 . The method of  claim 11 , further comprising merging the compact cluster representation based on at least one selection from the group consisting of a relevance-weight factor, a seller factor, a price factor, a category factor, and an image factor. 
     
     
         17 . The method of  claim 11 , further comprising calculating a relevance score using a demand factor. 
     
     
         18 . The method of  claim 17 , wherein the relevance score is calculated, in part, based on a comparison of a similarity of a demand category histogram and a supply category histogram. 
     
     
         19 . The method of  claim 17 , further comprising using the relevance score to prune irrelevant clusters. 
     
     
         20 . A system comprising:
 a hardware processor;   a retrieving engine to retrieve, from a database, item data;   a cluster generator to construct at least one base cluster having at least one document with common item data stored in a suffix ordering; and   a compacting engine to compact, using the hardware processor, the at least one base cluster to create a compacted cluster representation having a reduced duplicate suffix ordering amongst clusters.   
     
     
         21 . The system of  claim 21 , further comprising a merging engine to merge the compacted cluster representation to generate a merged cluster based upon a first overlap value applied to the at least one document with the common item data.