IBM's pioneering text mining research effort honored in Japan

In 1997, a team of researchers at IBM Research - Tokyo invented TAKMI, a technology that can read and uncover trends from the avalanche of information in natural language format. The Ministry of Education, Culture, Sports, Science and Technology of Japan recently honored the research team for its contribution in pioneering text mining technology with the 2012 Commendation for Science and Technology.

TAKMI (Text Analysis and Knowledge Mining) is a text mining technology that goes beyond search -- analyzing data from structured and numerical, to unstructured and text-based. It looks for unknowns by mining data such as email, product reviews on the Internet, memos, and other written documents.

“[What] unstructured information can tell you is the answer to questions you didn’t even know you needed to worry about. It lets you know what you don’t know,” said Scott Spangler of IBM Research - Almaden and co-author of Mining the Talk: Unlocking the Business Value in Unstructured Information.

Award Recipient (From left) Tetsuya Nasukawa,
Kohichi Takeda, Seiji Hamada,
Hiroshi Kanayama and Hideo Watanabe.
TAKMI also incorporates grammatical relationships into its analysis. Analyzing the Japanese language was a challenge for the research team because it does not contain white spaces as word separators, like English. The researchers used a natural language processing technique called dependency parsing that identifies which word is the subject, the verb, the object, and also examines the relationships between words. This technique was also used to help IBM Watson, the DeepQA system, learn natural language written in English. 

Today, the text mining technology pioneered by IBM Research is widely applied to industries including manufacturing, finance, insurance, broadcast, telecommunications and retail industries to help improve customer care, product and services quality, and expand business opportunities.  

Last year, the award was given to the IBM's accessibility research team led by IBM Fellow Chieko Asakawa in recognition of their contributions in the development of a voice browser for the visually impaired, which has since become the foundation for Web accessibility research and development, and for accessibility legislation and standardization around the world.

Labels: , , ,