Please use this identifier to cite or link to this item:
https://idr.l3.nitk.ac.in/jspui/handle/123456789/6981
Title: | Sigma encoded inverted files |
Authors: | Trotman, A. Subramanya, V. |
Issue Date: | 2007 |
Citation: | International Conference on Information and Knowledge Management, Proceedings, 2007, Vol., , pp.983-986 |
Abstract: | Compression of term frequency lists and very long document-id lists within an inverted file search engine are examined. Several compression schemes are compared including Elias ? and ? codes, Golomb Encoding, Variable Byte Encoding, and a class of word- based encoding schemes including Simple-9, Relative-10 and Carryover-12. It is shown that these compression methods are not well suited to compressing these kinds of lists of numbers. Of those tested, Carryover-12 is preferred because it is both effective at compression and fast at decompression. A novel technique, Sigma Encoding prior to compression, is proposed and tested. Sigma Encoding utilizes a parameterized dictionary to reduce the number of bits necessary to store an integer. This method shows an about 0.3 bit per integer improvement over Carryover-12 while costing only about 3 extra clock cycles per integer to decompress. Copyright 2007 ACM. |
URI: | http://idr.nitk.ac.in/jspui/handle/123456789/6981 |
Appears in Collections: | 2. Conference Papers |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.