Login
Computer Science @ Rochester
Tuesday, May 12, 2009
10:00 AM
Computer Studies Building, Room 703
Ph.D. Thesis Proposal
Mehdi Hafezi Manshadi
University of Rochester
Towards a Robust Deep Language Understanding System
Many applications of Natural Language Processing require deep understanding of text. Deep understanding is proved to be extremely hard. In order to overcome this difficulty, current NLP systems either simulate deep understanding by taking several shallow processing steps (part of speech tagging, word sense disambiguation, named entity recognition, syntactic parsing, semantic role labeling, etc.) or by restricting the domain of text (Database Query Applications, Air Travel Information Systems, Railroad Freight System, etc.).

In the first case, the system is not actually doing deep understanding, and in the second case the system is not robust as it is restricted to a domain-specific lexicon, grammar, ontology, etc. We first propose to develop a formalism, which allows one to represent both shallow and deep semantic analysis under a unified framework as the first step toward robustness. We then suggest using a deep understanding module as a core and leveraging some external resources to broaden the coverage of the system. External resources such as existing world knowledge resources, unlabeled data, etc. help to learn new information, which enables the system to understand new concepts and relations.