Knowledge Management & E-Learning: An International Journal (KM&EL), Vol 6, No 1 (2014)

Font Size:  Small  Medium  Large

Use of global context for handling noisy names in discussion texts of a homeopathy discussion forum

Mukta Majumder, Sujan Kumar Saha

Abstract


The task of identifying named entities from the discussion texts in Web forums faces the challenge of noisy names. As the names are often misspelled or abbreviated, the conventional techniques have failed to detect the noisy names properly. In this paper we propose a global context based framework for handling the noisy names. The framework is tested on a named entity recognition system designed to identify the names from the discussion texts in a homeopathy diagnosis discussion forum. The proposed global context-based framework is found to be effective in improving the accuracy of the named entity recognition system.


Full Text: PDF

Creative Commons License
This work is licensed under a Creative Commons Attribution 3.0 License.

Knowledge Management & E-Learning: An International Journal (KM&EL)
ISSN 2073-7904

 

Maintained and Developed by:

Laboratory for Knowledge Management & E-Learning

Faculty of Education, The University of Hong Kong