This Project's

Arabic Language Processing

Theme: Natural Language Processing
We were one of the pioneers using corpus based and shallow methods in Arabic language processing. We developed a cluster based information retrieval method, which uses clusters of words sharing a stem to index documents. We provided the first systematic approach to detecting broken plurals, and showed the impact of including broken plurals in retrieval. We released our corpus - the second available worldwide - through ELRA. We are part of the NEMLAR network.
Anne De Roeck Bashar Nuseibeh