Implementation of Conditional Random Field for Named Entity Recognition in Indonesian Traditional Arts Digital Article

— Digital articles are one type of media used as a reference when looking for information, one of which is information related to traditional arts in Indonesia. In traditional art articles, there are various artistic entities, such as dance, performing arts, music, artistic figures, and musical instruments. There are so many types of art in Indonesia, especially on the island of Java, making it more difficult to identify the entities in the article. This study aims to design a system that can label each entity in traditional Indonesian art articles through Named Entity Recognition (NER) using the Conditional Random Field (CRF) algorithm. The implementation of CRF in NER is carried out using the Python programming language, and is evaluated through the f1-score. Based on several test scenarios, the best performance was found in the distribution of training data and test data with a ratio of 4:1, which resulted in an average f1-score of 88.2%, recall 90.3%, and precision 88.2%.