Urdu Named Entity Recognition System Using Deep Learning Approaches

Abstract

Named entity recognition (NER) is a fundamental part of other natural language processing tasks such as information retrieval, question answering systems and machine translation. Progress and success have already been achieved in research on the English NER systems. However, the Urdu NER system is still in its infancy due to the complexity and morphological richness of the Urdu language. Existing Urdu NER systems are highly dependent on manual feature engineering and word embedding to capture similarity. Their performance lags if the words are previously unknown or infrequent. The feature-based models suffer from complicated feature engineering and are often highly reliant on external resources. To overcome these limitations in this study, we present several deep neural approaches that automatically learn features from the data and eliminate manual feature engineering. Our extension involved convolutional neural network to extract character-level features and combine them with word embedding to handle out-of-vocabulary words. The study also presents a tweets dataset in Urdu, annotated manually for five named entity classes. The effectiveness of the deep learning approaches is demonstrated on four benchmarks datasets. The proposed method demonstrates notable progress upon current state-of-the-art NER approaches in Urdu. The results show an improvement of 6.26% in the F1 score.

A deep learning approach for Named Entity Recognition in Urdu language

“…Determining the optimal n-gram window size depends on multiple factors including the nature of the text in a corpus, different sentence structures that may produce different patterns and dependencies between words, and target language and writing styles. For the Urdu language, various studies generally use an n-gram window size ranging from 3 to 5 for different NLP tasks while existing studies on Urdu NER [29][30][31]53] have used a window size of five. For this study, we tested different n-gram window size values ranging from three to five and selected the best-performing n-gram window size of five.…”

“…Deep learning (DL) based studies use different pre-trained word embedding techniques [21][22][23][24][25][26][27] to map the words in vectors using the language vocabulary to automatically extract meaningful relationships among words in the dataset. Due to limited vocabulary size, out-of-vocabulary words pose significant challenges for morphology-rich languages [28] like Urdu due to language complexities [29]. Existing studies have used DL models for Urdu NER [29][30][31] by utilizing Word2Vec, Txt2Vec, MKW2v, and GloVe embeddings.…”

“…Due to limited vocabulary size, out-of-vocabulary words pose significant challenges for morphology-rich languages [28] like Urdu due to language complexities [29]. Existing studies have used DL models for Urdu NER [29][30][31] by utilizing Word2Vec, Txt2Vec, MKW2v, and GloVe embeddings. The hybrid approach [32] uses a combination of rule-based and ML and DL approaches that perform well in domain-specific scenarios.…”

Named Entity Recognition Using Conditional Random Fields

“…Supervised systems infer rules through pre-labeled input, referred to as training data, and conform to estimation methods, nonparametric, or kernel-based learning algorithms, as well as logic-based algorithms [23,25]. NER frameworks based on machine learning are often more customizable than rule-based techniques [26]. An ML technique could adjust to new contexts with little cost provided that training data is freely available [18].…”