Exploring Computational Linguistics for Metadata Building (CLiMB) Judith Klavans Deputy Vice Presdient Columbia University Stephen Paul Davis Director of Library Systems Columbia University Angela Giral Columbia University The newly-initiated Computational Linguistics for Metadata Building (CLiMB) project at Columbia University aims to explore the use of natural language processing to automatically assist in the idenfication and extraction of meaningful descriptive metadata from texts associated with image collections. The collections we are using are architectural drawings from the Greene & Greene brothers, a set of lithographs of Chinese Gods on paper, and images from the Digital South Asia Library collection (dsal). This is a newly initiated project, and is experimental in nature. We will present the techniques we are exploring, our goals, and progress. Web Links: http://www.columbia.edu/cu/cria/climb/