Retrieval of XML data - to support NLP applications

摘要

Information retrieval (IR) deals with the organization, storage, representation and access to information. An XML document is a database only in the strictest sense of the term. That it is a collection of data. As a database XML has some advantages like it is self-describing, portable & it can describe data in tree or graph structures. At the same time it is considered that XML is verbose and access to data is slow due to parsing & text conversion. Even though it is tough to provide efficient storage, indexes, security, transactions & data integrity, multi-user access, triggers, queries across multiple documents etc. with XML data repository, the application of XML in present & next generation applications and the huge amount of XML data available round the world can not be ignored. So retrieving XML data efficiently - is always a challenge. Hybrid Data Servers, SQL, XPath and WQuery have opened a new era to handle both relational and XML data. XML, has also become the standard framework for publishing on the net, as well as the standard e-commerce language to build B2B and B2C Web services. A major concern for this scenario is the "point of creation" bottleneck, at which creating useful, well-structured XML data can consume unduly amount of time and effort. , XML helps the NLP researches, especially the ones with annotated corpus based approaches, by providing them with the knowledge representation frameworks for morphological, syntactic, semantics and/or pragmatics information structure of NL resources. In many cases, XML is able to provide NLP with deeper semantic structure clues and thus realize much more robust, higher precision NLP applications. This paper focuses on the retrieval techniques of XML databases for present and next generation NLP applications.