INCA (INtelligent Semantic CAching)
Project @ UCLA

since 09/20/1999

Gold Hand

Around the year 1200 A.D. the Incas, came down from high in the Andes mountains to conquer the scattered tribes throughout the South American continent. They imposed their rule and their highly organized form of government on the weaker, less organized peoples of what is now Ecuador, Peru, Bolivia and parts of Chile. The agricultural and organizational skills they brought down with them served them well. Within 100 years they had built a powerful Empire that spanned from northern Ecuador to Chile. They possessed great skills in masonry, as their magnificent stone walls and buildings show. (taken from here)



Welcome to the INCA project home page! We, CoBase research group, aim at building a novel semantic caching mechanism that is more intelligent that the conventional ones.

Semantic caching exploits the semantic locality of the queries by caching a set of semantically associated results, instead of tuples or pages which are used in conventional caching. The semantic caching can be particularly effective in improving performance when a series of semantically associated queries are asked, thus the results may likely overlap orcontain one another.

Research Areas

Semantic Caching, View Materialization, Query Optimization

Overview

Machu Picchu We present a new semantic caching scheme suitable for web sources. Since the web sources have typically weaker querying capabilities than conventional databases, it is not trivial to apply existing semantic caching schemes directly. We provide a seamlessly integrated query translation and capability mapping between the wrappers and web sources in the semantic caching to cope with such difficulties and describe several related issues. In addition, an analysis on the match types between the user's input query and queries stored in the cache is presented. We show how to use semantic knowledge acquired from the data to avoid unnecessary access to the web sources by transforming the cache miss to the cache hit. Further, a polynomial time algorithm based on the extended and knowledge-based matching is proposed to find the best matched query in the cache. Finally, experimental results are presented to illustrate the effectiveness of our proposed semantic caching scheme and an area of application of the proposed technique is given.

Member

Wesley W. Chu : Principal Investigator
Dongwon Lee : Ph.D Student

Publications


[Note: Photo of the Machu Picchu was taken from here]