MontyLingua (and ConceptNet) to simplify natural text processing tasks

By:

Natural text processing is generally a difficult task due to the imprecision of human language and its dependence on epistemological knowledge, which includes common sense and cultural experiences. However, natural text processing, a subset of natural language processing, has a large impact on information and knowledge management as most of human knowledge exists in unstructured human texts rather than databases. In this talk, I will demonstrate how MontyLingua, a text processor written in Python, can be used for common text processing tasks, and give a brief comparison of MontyLingua and other natural text processing libraries, namely GATE (Java-based) and NLTK (Python-based). I will also demonstrate some uses of ConceptNet (http://www.openmind.org/commonsense), an epistemological database incorporating MontyLingua, in common natural text processing tasks, such as document indexing, basic information extraction, and emotional estimation.

Stream: Python

Paper: A paper has not yet been submitted.

Maurice Ling is a PhD candidate in the department of Zoology of the University of Melbourne working on text analysis of biological literature for the purpose of understanding hormone interactions in the mouse mammary cell.