This repository is a supplement resource for a research article entitled "Deep Learning Untuk Entity Matching Produk Kamera Antar Online Store Menggunakan DeepMatcher". The research is conducting deep learning for entity matching using DeepMatcher by Anhaidgroup. The test is performed on Google Colaboratory which can be accessed on this link.
In Computer Science field, Entity Matching has become a challenge for some researchers. Some have tried to develop an entity matching algorithm to improve accuracy. This study will test DeepMatcher as a representation of Entity Matching using Deep Learning by matching entities against case studies of matching camera products at two online stores using four different learning algorithms that DeepMatcher has, namely Smooth Inverse Frequency, Bidirectional RNN, Decomposable Attention Model, and Hybrid. Model. By building a dataset and a learning model, DeepMatcher can independently match data that has not been previously entered. The matching results will be measured using an f-measure to then analyze its reliability. The test results show that the type of learning on DeepMatcher that is most suitable for using in entity matching on camera products between online stores is Bidirectional RNN with an average resulting F1 score of 61.546.
Email - zainadam.id@gmail.com
Twitter - @sensasi_DELIGHT