New techniques for compressing and storing inverted lists are presented. Differently from previous research, these techniques are especially designed for volatile inverted lists and combine different types of compression (including prefix compression) with block segmentation to allow easy insertion/deletion of pointers and, most importantly, to significantly reduce execution times while keeping storage requirements close to a baseline monolithic inverted list implementation based on Elias’s ( codes. Inverted lists for information retrieval are addressed and experiments are reported. The best method uses an optimized block-oriented evaluation that is able to efficiently skip irrelevant pointers and that has an observed average execution time which is less than 65% of the baseline implementation.
Fast Block-Compressed Inverted Lists
SACCO, Giovanni
2012-01-01
Abstract
New techniques for compressing and storing inverted lists are presented. Differently from previous research, these techniques are especially designed for volatile inverted lists and combine different types of compression (including prefix compression) with block segmentation to allow easy insertion/deletion of pointers and, most importantly, to significantly reduce execution times while keeping storage requirements close to a baseline monolithic inverted list implementation based on Elias’s ( codes. Inverted lists for information retrieval are addressed and experiments are reported. The best method uses an optimized block-oriented evaluation that is able to efficiently skip irrelevant pointers and that has an observed average execution time which is less than 65% of the baseline implementation.File | Dimensione | Formato | |
---|---|---|---|
dexa2012.pdf
Accesso riservato
Tipo di file:
POSTPRINT (VERSIONE FINALE DELL’AUTORE)
Dimensione
206.45 kB
Formato
Adobe PDF
|
206.45 kB | Adobe PDF | Visualizza/Apri Richiedi una copia |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.