A research paper is one of the types of scientific works created by academics to demonstrate their level of expertise in a field. Since a research paper is a scientific work, its writing must be accurate in terms of knowledge and language. Errors in writing scientific works, especially research papers, often occur, and sometimes due to the large number of words in the paper, authors often have difficulty finding writing errors. Based on this problem, an application is needed to find writing errors in a research paper. In searching for incorrect words, a word matching process will be performed between the paper and a dataset, where the dataset will be in JSON format to facilitate the search process. After matching the words from the document and the dataset, words that are different from those in the dataset will be obtained, and then a matching process between the incorrect words and the words in the dataset will be conducted using the Jaro-Winkler Distance method to match the words and evaluate their similarity. The words in the dataset will be derived from the KBBI (Great Dictionary of the Indonesian Language) as a reference for Indonesian words. The application to be developed will use the Android Studio software, which is useful for creating mobile-based applications. The Jaro-Winkler Distance method has a high level of accuracy and precision, around 80.6%. This level of accuracy and precision indicates that the Jaro-Winkler Distance method is quite suitable for use in providing suggestions for incorrectly spelled words or typographical errors.
Copyrights © 2024