The purpose of this talk is to introduce researchers and students at Maryland to Microsoft's Natural Language Software Development Kit (SDK), and to demonstrate the type of feedback Microsoft hopes to solicit from them, as part of a licensing agreement between Microsoft and Maryland, for the English version of the SDK as part of an project funded by an NSF grant to Philip Resnik (University of Maryland) and Christiane Fellbaum (Princeton University).
I present a project developed this spring by an intern at Microsoft, building software on an internally-relased version of Microsoft's Natural Language Software Development Kit (SDK). The intern, Jiang Jiang Chen, is a 2002 computer science graduate of the University of Washington, entering graduate school in computer science at Stanford this fall. The purpose of her project was to dogfood a multilingual version of the SDK as a client who is computationally savvy but linguistically untrained. The application is a semantic transfer system for German to English translation, incorporating a bilingual lexicon, and rules written by the developer to transfer between semantic structures. The talk will outline what was required to develop this a prototype application, as well as the kind of problems she encountered.
Mari Broman Olsen is a lexical semanticist in the Natural Language Group at Microsoft Corporation.
For the colloquium series schedule, see the UMD Computational Linguistics Colloquium Series web page at http://umiacs.umd.edu/~resnik/cl_colloquium/. If you are interested in meeting with the speaker, please contact Philip Resnik (resnik@umiacs.umd.edu).