Object recognizer for organizing and structuring unstructured documents using Interrogative Knowledge
Iskandar Ishak, Fatimah Sidi, Marzanah A. Jabar · Universiti Putra Malaysia Institutional Repository (Universiti Putra Malaysia) · 2016
Knowledge in unstructured documents is lacking the structured characteristics; therefore raising the problem of extracting answers when queried by simple queries. This paper proposes the Interrogative Knowledge Object- (IKO-) Recognizer that is able to model the extracted interrogative lexical constructs from unstructured documents into ontological constructs. An experiment is carried out to test performance of the IKO-Recognizer using quantitative retrieval performance metrics of recall and precision. The result shows that the extracted interrogative constructs is able to increase human understanding on knowledge available in unstructured document.