Neural POS tagging of shahmukhi by using contextualized word representations

Tehseen, Amina; Ehsan, Toqeer; Liaqat, Hannan Bin; Ali, Amjad; Al‐Fuqaha, Ala

doi:10.1016/j.jksuci.2022.12.004

Cited by 7 publications

(4 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Beside the contextualized word embeddings, we also experimented by incorporating part of speech(POS) tags and Word2Vec embeddings. POS tagging has not shown any improvements for NER (Tehseen et al, 2022(Tehseen et al, , 2023 and the F 1 -score remained around ∼0.78. We used the Stanford POS tagger (Toutanova et al, 2003) to tag the Wojood NER dataset and concatenated the POS encoding vectors with the word encoding vectors at the input layers of the models.…”

Section: Modelsmentioning

confidence: 88%

“…Different machine and deep learning techniques have been used to perform NER, such as, Conditional Random Fields (CRF) (Patil et al, 2020;Bhumireddypalli et al, 2023), Support Vector Machines (SVM) (Mady et al, 2022), template-based (Cui et al, 2021), Recurrent Neural Networks (RNN) (Ahmad et al, 2020), Bidirectional LSTM (Tehseen et al, 2023), Transformer-based Models (e.g. BERT) Agrawal et al, 2022) and others.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

AlphaBrains at WojoodNER shared task: Arabic Named Entity Recognition by Using Character-based Context-Sensitive Word Representations

Ehsan,

Ali,

Al-Fuqaha

2023

Proceedings of ArabicNLP 2023

View full text Add to dashboard Cite

This paper presents Arabic named entity recognition models by employing single-task and multi-task learning paradigms. The models were developed by using character-based contextualized Embeddings from Language Model (ELMo) in the input layers of the Bidirectional Long-Short Term Memory (BiLSTM) networks. The ELMo embeddings are quite capable of learning the morphology and contextual information of tokens in word sequences. The single-task learning model outperformed the multi-task learning model, achieving micro F 1 -scores of 0.8751 and 0.8884, respectively, ranking 10 th and 7 th in the shared task for flat and nested NER.

show abstract

Section: Modelsmentioning

confidence: 88%

Section: Introductionmentioning

confidence: 99%

AlphaBrains at WojoodNER shared task: Arabic Named Entity Recognition by Using Character-based Context-Sensitive Word Representations

Ehsan,

Ali,

Al-Fuqaha

2023

Proceedings of ArabicNLP 2023

View full text Add to dashboard Cite

show abstract

“…They used the available Malayalam corpus of 280000 tokens and trained different RNN models including GRU, LSTM and Bi-LSTM that shows better results than the available POS taggers reaching 98% in terms of F-measure. Also, a bidirectional LSTM POS tagger for Shahmukhi is proposed in the study [20]. They first collected millions of tokens and manually annotated 130000 words to be tested in the study.…”

Section: Literature Review In Pos Tagging Low Resourced Language Usin...mentioning

confidence: 99%

Deep Neural Networks for Part-of-Speech Tagging in Under-Resourced Amazigh

Bani¹,

Amri²,

Zenkouar³

et al. 2023

RIA

View full text Add to dashboard Cite

Part-of-speech (POS) tagging denotes the assignment of appropriate grammatical categories to individual words within a sentence or text, playing a pivotal role in numerous natural language processing (NLP) tasks. While POS tagging in widely-used languages such as English has reached accuracy levels exceeding 97%, less-resourced languages such as Amazigh have seen limited research and therefore, less accuracy in tagging efforts. This paper aims to bridge this gap by exploring the application of deep learning models for Amazigh POS tagging, specifically focusing on various types of recurrent neural networks (RNNs)-gated recurrent networks, long short-term memory (LSTM) networks, and bidirectional LSTM networks. Despite the relatively small dataset of 60k tokens, a stark contrast to the vast corpuses available for languages with extensive resources, the proposed RNN models have demonstrated significant improvements over existing Amazigh POS taggers. Remarkably, all RNN models tested in this study outperformed traditional machine learning taggers, achieving an accuracy rate of 97%, thus presenting a promising avenue for enhanced POS tagging in under-resourced languages. This research underscores the potential of deep learning approaches in contributing to the advancement of linguistic studies in less-documented languages, such as Amazigh.

show abstract

“…POS tagging, also known as grammatical tagging is the process of labeling a word in a sentence as relating to a part of speech, based on both its definition and its context (1) . POS tagging is employed as a preprocessing activity by other Natural language processing (NLP) applications like machine translation and relation extraction to enhance their performance, it is a crucial study area to achieve high-quality research https://www.indjst.org/ in other NLP areas (2) . An example of POS tagging is given below, For example, <PPR> <JINT><AMN><<VM><VA><PUN>/Tai bohut polom koi bhabi pale/She understood very late.…”

Section: Introductionmentioning

confidence: 99%

Parts-of-Speech Tagging for Unknown Words in Assamese using Viterbi Algorithm

Phukan,

Baruah,

Sarma

et al. 2023

IJST

View full text Add to dashboard Cite

Objectives: This study aims to explore the application of the Viterbi algorithm for Part-of-Speech (POS) tagging in the Assamese language, focusing on tagging out-of-vocabulary words. The objective of this paper is to assess the algorithm's performance using various training and testing data ratios of 50:50, 70:30, and 90:10. Methods: The study utilizes the dynamic programming capabilities of the Viterbi algorithm to determine the most likely sequence of hidden states based on observable events. A corpus comprising approximately 50,000 words is employed to train the algorithm, with different ratios of this data utilized for training and testing purposes. Findings: The Viterbi algorithm achieves an accuracy of 86.34%, surpassing the state-of-the-art POS taggers for the Assamese language. The experimental evaluation demonstrates that the proposed approach outperforms previously existing research work by achieving 6.14% higher accuracy in tagging out-of-vocabulary words, highlighting its effectiveness in addressing the challenges associated with less-resourced languages like Assamese. Novelty: The results of this study contribute to the understanding and development of POS tagging techniques in less-resourced languages like Assamese. The proposed approach not only achieves superior performance in terms of accuracy but also showcases its potential for improving POS tagging in similar linguistic contexts, surpassing the achievements of previous research efforts.

show abstract

Neural POS tagging of shahmukhi by using contextualized word representations

Cited by 7 publications

References 30 publications

AlphaBrains at WojoodNER shared task: Arabic Named Entity Recognition by Using Character-based Context-Sensitive Word Representations

AlphaBrains at WojoodNER shared task: Arabic Named Entity Recognition by Using Character-based Context-Sensitive Word Representations

Deep Neural Networks for Part-of-Speech Tagging in Under-Resourced Amazigh

Parts-of-Speech Tagging for Unknown Words in Assamese using Viterbi Algorithm

Contact Info

Product

Resources

About