Department of Computer Science at UH

University of Houston

Department of Computer Science

In Partial Fulfillment of the Requirements for the Degree of
Master of Science

Nathan Leach

Will defend his thesis

Modeling the Recognition and Understanding of Time Terms in the ALPS System


When humans communicate, we often have to set a temporal context for the conversation as part of the communication process. The use of this temporal context allows the human to convey the idea that something either had happened in the past, is currently happening, or will happen in the future. It may also be used to indicate the exact timing of an action or the relative ordering of different events. The setting of this temporal context, however, is not static; that is, there is often no point where a conversation stops and all participants mutually agree to the temporal context to use for the next part of the conversation. Instead, the temporal context is established using many explicit and implicit terms for time during the course of the conversation.

While this temporal context is easy for humans to understand, the subtleties in the meanings of the time terms we are communicating make it somewhat difficult to model this same understanding as a computer program. This thesis describes the beginning of modeling this understanding of time terms as a computer program. The model begins with the ability to understand time adverbs by learning knowledge about the meaning of these adverbs. The model expands with capabilities to combine adverbs and time concept values as they are used in a sentence separated by prepositions such as at, on, or in to express a time concept value. Finally, the model provides validation capabilities that allow for the detection of incorrect time values. This validation capability is facilitated by the ability to consult with external entities that contain expert knowledge about the subject of time.

Date: Wednesday, April 27, 2011
Time: 11:00 AM
Place: 550-PGH
Faculty, students, and the general public are invited.
Advisor: Dr. Kam-Hoi Cheng