Know-How Extraction from Textual Data for Problem Solving

P. Soison and C. Pechsiri (Thailand)


Elementary discourse unit, Know-how, Proceduralknowledge boundary, Target


To extract the procedural knowledge or Know-how from unstructured textual data is the challenging work. This paper presents how to automatically extract all Know how from technical documents on the website for solving the same health problems. The research extracts the informative Know-how in the form of multiple EDUs (Elementary Discourse Units) of herbal-medicine preparation. There are two problems of extracting the procedural knowledge, the identification problem and the boundary determination problem. We propose using Naive Bayes by sliding the window size of two consecutive EDUs with the sliding distance of one EDU to resolve the extraction problems after knowing the target of health solutions. The results from our proposed methodology are 85% precision and 76 % recall.

Important Links:

Go Back