The HAIRCUT System at TREC-9

The Hopkins Automated Information Retriever for Combing Unstructured Text (HAIRCUT) is a research IR system developed at the Johns Hopkins University Applied Physics Laboratory (JHU/APL). HAIRCUT benefits from a basic design decision to support flexibilit y throughout the system. One specific example of this is the way we represent documents and queries; words, stemmed words, character n-grams, multiword phrases are all supported as indexing terms. This year we concentrated our efforts on two of the tasks in TREC-9, the main web task and cross-language retrieval in Chinese and English.