??? PMaps are beneficial when there is limited space for index creation, as they
occupy much less space compared to bitmaps indexing the same set of attributes.
??? PMaps perform well for multi-attribute queries, even better than single attribute
queries, unlike REBSIs and many other known techniques. Hence, it is
useful to create PMaps for frequently used multi-attribute queries.
??? PMap savings increase in the case of large databases. Therefore, PMaps could
be created for very large databases where SBIs are inefficient. PMap savings
also increase for larger block sizes when the database size is large.
??? PMaps perform very well in the case of inequality queries or high selectivity
queries. This result could be used to create PMaps for specific applications.
Extensions to the research reported here are summarized as follows:
??? Currently, PMaps are not designed to represent aggregation and grouping
queries. Grouping attributes are usually low cardinality attributes, and can be
covered by enumerated properties in a PMap. After processing the WHERE
clause, the pstrings of the tuples in the intermediate result need to be scanned
for the ones satisfying the grouping property. Future research could extend
the PMap to solve aggregation and grouping queries.
??? Index maintenance, including deletions and updates, is an area for future investigation.
Pages:
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386