In the original Hadith Document, which is in Arabic, there are inconsistencies in mention the name of the narrator in the Sanad, such as inconsistencies in using the full name or the nickname. In addition, other problem arise when the hadith is translated into Indonesian to make it easier for Indonesian who don't understand Arabic, such as inconsistencies during translation and typographical errors. In general, the problem that occurs is that there are the different mention of names in hadith document but can refer to the same entity. As a solution, the name matching method is needed that can match the mention to the referenced entity. This study is important to be able to recognize the entities of hadith narrator well for people who learn the science of hadith. In previous study, Muazzam did grouping the mention of hadith narrators based on the text, but not on the entity. This study will focus on matching the names of narrator with different mention but refer to the same entity using the neural network-based methods. The Simple Neural Network (Simple NN) can capture information based on experience from the data learned, namely data on the names of hadith narrators. The input is the mention name of the selected narrator with its features. The Simple NN model will learn the data, then the model can match the selected mention to the original entity. This study can reduce the gap from previous study, so that the name of narrator can be refer to the original entity. This model successfully to identify the original entity from the mention of narrator with a 91.30% of accuracy.
Keywords: Arabic Name, Name Entity, Name Matching, Name of Hadith Narrators, Simple Neural Network