Package org.apache.mahout.cf.taste.hadoop.similarity.item

Class Summary
CountUsersKeyWritable a writable key that is used by CountUsersMapper and CountUsersReducer to count unique users by sending all userIDs to the same reducer and have them sorted in ascending order so that there's no buffering necessary when counting them
CountUsersKeyWritable.CountUsersGroupComparator all userIDs go to the same reducer
CountUsersKeyWritable.CountUsersPartitioner all userIDs go to the same partition
CountUsersMapper Maps out the userIDs in a way that we can use a secondary sort on them
CountUsersReducer counts all unique users, we ensure that we see userIDs sorted in ascending order via secondary sort, so we don't have to buffer all of them
ItemSimilarityJob Distributed precomputation of the item-item-similarities for Itembased Collaborative Filtering
MostSimilarItemPairsMapper  
MostSimilarItemPairsReducer  
ToItemVectorsReducer  
 



Copyright © 2008-2010 The Apache Software Foundation. All Rights Reserved.