Bidirectional Recurrent Neural Network Approach for Arabic Named Entity Recognition

Mohammed Ali,Guanzheng Tan,Aamir Hussain

doi:10.3390/fi10120123

Mohammed Ali, Guanzheng Tan + Show 1 more

Open Access

https://doi.org/10.3390/fi10120123

Copy DOI

Journal: Future Internet	Publication Date: Dec 13, 2018
Citations: 25	License type: CC BY 4.0

Affiliation: Central South University

Abstract

Recurrent neural network (RNN) has achieved remarkable success in sequence labeling tasks with memory requirement. RNN can remember previous information of a sequence and can thus be used to solve natural language processing (NLP) tasks. Named entity recognition (NER) is a common task of NLP and can be considered a classification problem. We propose a bidirectional long short-term memory (LSTM) model for this entity recognition task of the Arabic text. The LSTM network can process sequences and relate to each part of it, which makes it useful for the NER task. Moreover, we use pre-trained word embedding to train the inputs that are fed into the LSTM network. The proposed model is evaluated on a popular dataset called “ANERcorp.” Experimental results show that the model with word embedding achieves a high F-score measure of approximately 88.01%.

Highlights

The Named entity recognition (NER) is important in natural language processing (NLP) tasks used to detect named entities (NEs) in texts and classify them into predefined categories, such as location, person, time, date, and organization [1]
The results showed that high accuracy on the same dataset (“ANERCorp”) is achieved when an artificial neural network (ANN) method is adapted to the NER system than when complemented with the decision trees
We experiment with a B-Recurrent neural network (RNN) with long short-term memory (LSTM)/Gated Recurrent Units (GRUs) for the Arabic NER (ANER) task

Summary

Introduction

The Named entity recognition (NER) is important in natural language processing (NLP) tasks used to detect named entities (NEs) in texts and classify them into predefined categories, such as location, person, time, date, and organization [1]. NER is a crucial preprocessing phase in various NLP applications to improve the overall performance. It extracts valuable information from raw data and simplifies downstream tasks, such as text clustering, information retrieval, translation, and question answering [2]. Arabic is a Semitic and the standard language spoken in the Arab world. In the Arab world, around 360 million people speak Arabic in more than 25 countries [4]

Results

Discussion

Conclusion