Applied Information Extraction technique for extracting the king name who build a temple in Lanna Kingdom from historical documents

Abstract

The objective of this study is to propose a technique to extract the king name in Lanna kingdom, who built the temple, from historical documents. The research processes began with collecting the data of temples in Amphor Muang Chiangmai, which is the capital and administrative center of the Lanna kingdom. The data transformed into electronic document or a text file are passed to the LexTo program for word segmentation with the nearly correct meaning. After that designing algorithm and the development of Java program to extract the king name who built the temple was proceeded. The algorithm is designed based on Named Entity Recognition (NER) by defining king names and some keywords, that indicates the temple building, in form of dictionary to be used as a reference. Then the experiment followed by the evaluation of the performance is set on 100 temples, by comparing the extracted results to the correct answers from the original documents. The accuracy is 97 percent. From the result indicates that the proposed technique can be efficiently applied to extract the king name of the temple builder in the Lanna kingdom from the historical documents.