Abstract: Provided is a method and computer program product for determining a document relevance function for estimating a relevance score of a document in a database with respect to a query. For each of a plurality of test queries, a respective set of result documents is collected. For each test query, a subset of the documents in the respective result set is selected, and a set of training relevance scores is assigned to documents in the subset. In one embodiment, at least some of the training relevance scores are assigned by human subjects who determine individual relevance scores for submitted documents with respect to the corresponding queries. Finally, a relevance function is determined based on the plurality of test queries, the subsets of documents, and the sets of training relevance scores.

Abstract: Provided is a method and computer program product for determining a document relevance function for estimating a relevance score of a document in a database with respect to a query. For each of a plurality of test queries, a respective set of result documents is collected. For each test query, a subset of the documents in the respective result set is selected, and a set of training relevance scores is assigned to documents in the subset. In one embodiment, at least some of the training relevance scores are assigned by human subjects who determine individual relevance scores for submitted documents with respect to the corresponding queries. Finally, a relevance function is determined based on the plurality of test queries, the subsets of documents, and the sets of training relevance scores.

Abstract: An embodiment of the present invention provides a method and apparatus for sorting very large data sets using a parallel merge sort. Given sorted work files S1, . . . , Sp, produced by P processes, the described embodiment of the method effectively implements a parallel merge onto respective output partitions O1, . . . , Op of the processes P. Because each of these output partitions O has a finite size, the invention must quickly determine “splitting keys” for each output partition O in such a way that the data in the work files will be split between the multiple output partitions O without overrunning the size of any of the partitions O. Once the splitting keys for each partition are determined, the processes exchange data so that the output partitions of each process contains data between the splitting keys associated with that output partition.

Abstract: A method and apparatus that determines and uses two nearly uniform independent hash functions. The hash functions are created using only linear arithmetic and 4-byte machine register operations and, thus, can be created very quickly.

Abstract: To perform a sort of N records, a two-dimensional tree structure is formed with a tree of subtrees, where each subtree is formed by a plurality of nodes organized in a binary tree. For each leaf node in the tree, there is an ancestor chain of nodes (from child to parent, from parent to grand-parent, from grand-parent to great-grand-parent, . . . and so on) that connects each leaf node to the root node. To perform the sort, the processing unit stores codes representing keys into nodes in the two-dimensional tree and performs a tree sort of the keys using the codes. The codes are accessed in the subtrees and processed to determine the sort order of the keys and therefore the sort order of the corresponding records.