PENERAPAN SEMANTIC GRAPH DALAM PENGEMBANGAN PERANGKAT LUNAK UNTUK PERANKINGAN BERITA
Daftar Isi:
- The inefficiency of traditional news searching method causes the low of accuracy level in news searching its result so that the displays news is not relevant to the given input. Therefore, a system that can display the rankings of news is required to search related information with the input among many new news contained in the database. The ranking order of Indonesian news can be done by applying Semantic Graph method. News data used in this study is 179 updated Indonesian news that taken from 2 online news sites. The main process in this study is Pre- Processing, finding word synet in WordNet dictionary, forming Semantic Graph, calculating Semantic Similarity of the formed graph, and ranking the results of Semantic Similarity calculations. Pre-processing step consists of case folding, tokenizing, stopword removal processing, and case folding and tokenizing on keyword input processing. The relevance level of this study was measured using recall, precision, and based on user ratings. This research succeeded to get the relevance value of recall average 80% after updating data from previous data is 77,23%. The precision value after updating data tends to decrease to 20% from the previous range of 23.32%. User rating after updating data news is about 70% from 63%.