5.2 PF/Tijah Text Indexing

The database group of University Twente has created the Tijah XML retrieval system. In PF/Tijah, it has become fully integrated and distributed with MonetDB/XQuery.

You can find documentation at the PF/Tijah website. This site also provides information on how to create text indices on XML documents stored in MonetDB/XQuery and how to perform keyword search on them.

NOTE: it is currently not possible to update XML documents that have a text index on them. It is possible though, to add and delete entire XML documents from text-indexed document collections. If you have an application that queries a large XML document collection, while at the same time new documents are added frequently (or deleted), it is still a good idea to store your data in an updatable collection (even while it is not allowed to actually use in XQuery updates). The reason is that index maintenance is cheaper on updatable collections than on read-only collections (see Performance Tips).

PF/Tijah is highly configurable in terms of the ranking functions supported. Additional info: the OSIR paper provides a quick technical background on PF/Tijah.