Merging Schemas in a Collaborative Faceted Classification System

We have developed a system that improves access to a large, growing image collection by allowing users to collaboratively build a global faceted (multi-perspective) classification schema. We are extending our system to support both global and local schemas, where global schema provides a complete and uniform view of the collection whereas local schema provides a personal, possibly incomplete and idiosyncratic view of the collection. We argue that although users usually focus on their personal schemas, it is still desirable to have a global schema for the entire collection even if such local schemas are available. In order to keep the global schema updated in time. there is a need to develop a way to merge latest local schemas into the global one. This study focuses on such schema merging problem. To the best of our knowledge, it is the first attempt to solve the problem. We present a general standard of characterizing good schema merging algorithms and argue that three types of information, i.e., document content information, document categorization information and category linguistic information, must be considered in a good schema merging algorithm. We discuss the feasibility of applying existing data integration methods to solve the schema merging problem and provide ways to turn these methods into schema merging algorithms. We also propose a new, potentially good schema merging algorithm