A Natural Language Processing Framework for a Balinese Story Digital Portal

Funding period : 2022- Deactivate

Abstrak

Stories in Balinese are one of the cultural heritages that are loaded with local wisdom. These stories were often passed from one generation to the next orally. The development of information technology coupled with easier internet access has opened up opportunities for the preservation of these Balinese stories into a digital portal. To further attract the interest of the younger generation, this digital portal can be equipped with modern features developed with natural language processing. These features include automatic story clustering based on topic, summarizing each story using extractive methods and optimizing feature weights using genetic algorithms, identifying characters in each story using named entity recognition, classification of each character and a word-based story search feature. key. These features are integrated into a framework and are applied automatically to the collection of Balinese stories used. This research has succeeded in developing and implementing a Balinese language processing framework on the server side so that clustering, summarizing, character identification and classification features in stories can be done automatically. The developed digital portal interface can be accessed online.