Relation Detection for Indonesian Language using Deep Neural Network -- Support Vector Machine

TL;DR

This paper explores relation detection between Indonesian language entities using a neural network approach combined with SVM, achieving an F1-score of 0.8083 with a hybrid model.

Contribution

It introduces a hybrid neural network and SVM model for relation detection in Indonesian, utilizing multiple embeddings and hyperparameter tuning.

Findings

Best F1-score of 0.8083 achieved with CNN and SVM

Hybrid model outperforms individual classifiers

Effective use of multiple embeddings for relation detection

Abstract

Relation Detection is a task to determine whether two entities are related or not. In this paper, we employ neural network to do relation detection between two named entities for Indonesian Language. We used feature such as word embedding, position embedding, POS-Tag embedding, and character embedding. For the model, we divide the model into two parts: Front-part classifier (Convolutional layer or LSTM layer) and Back-part classifier (Dense layer or SVM). We did grid search method of neural network hyper parameter and SVM. We used 6000 Indonesian sentences for training process and 1,125 for testing. The best result is 0.8083 on F1-Score using Convolutional Layer as front-part and SVM as back-part.