GNoM: Graph Neural Network Enhanced Language Models for Disaster Related Multilingual Text Classification

Ghosh, Samujjwal; Maji, Subhadeep; Desarkar, Maunendra Sankar

doi:10.1145/3501247.3531561

Cited by 12 publications

(6 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The time complexity of different techniques in classifying the aspects present in the embedded documents is presented in Table 6. We compared the efficiency of our proposed methodology with different techniques such as support vector machine (SVM)‐ genetic algorithm(GA), 28 bidirectional encoder representations from transformers (BERT) ensemble model, 29 graph neural network based multilingual text classification framework (GNoM), 30 and Bert‐based mental approach 31 . The results show that the proposed model consumes less time when compared to the SVM‐GA and GNoM models.…”

Section: Resultsmentioning

confidence: 99%

Local search five‐element cycle optimized reLU‐BiLSTM for multilingual aspect‐based text classification

Kumar¹,

Sulochana

2022

Concurrency and Computation

View full text Add to dashboard Cite

Aspect-based sentiment analysis has gained wide popularity due to its benefits of text extraction, classification, and ranking the overall sentiments of each feature extracted. However, the aspect-based feature extraction techniques often result in acquiring more number aspects that refer to the same feature which arises the need for aspect-based text classification. Since most of the existing techniques focus on monolingual aspect-based sentimental analysis, we planned to develop a multilingual aspect-based text classification for Indian languages. We perform the multilingual aspect-based text classification on different morphologically rich and complex languages such as Hindi, Tamil, Malayalam, Bengali, Urdu, Telugu, and Sinhalese. To achieve this objective, in this article we present an optimized rectified linear unit (reLU) layer-based bidirectional long short-term memory (reLU-BiLSTM) deep learning tool is developed. The parameters of the reLU-BiLSTM architecture are optimized using the local search-based five-element cycle optimization algorithm (LSFECO) optimization algorithm. Initially, the proposed model preprocesses the multilingual texts obtained from the reviews using different techniques such as tokenization, special character removal, text normalization and so forth. The discrete and categorical features from the different languages are initially extracted by applying the bidirectional encoder representations from transformers (BERT) model which processes the sentences in the text in a layer-by-layer manner. The context learning and word embeddings (aspects) present in the text are identified using different approaches such as word mover's distance, continuous Bag-of-Words (CBOW), and Cosine similarity. The LSFECO optimized reLU-BiLSTM architecture classifies the different aspects present in the embedding document to its corresponding classes (flowers, plants, animals, sports, politics, etc). The efficiency of the proposed methodology is evaluated using the text obtained from different text documents such as semantic relations from Wikipedia, Habeas Corpus (HC) Corpora, Sentiment Lexicons for 81 Languages, IIT Bombay English-Hindi Parallel Corpus, and Indic Languages Multilingual Parallel Corpus. When compared to conventional techniques, the proposed methodology outperforms them in terms of entropy, coverage, purity, processing time, accuracy, F1-score, recall, and precision.

show abstract

Section: Resultsmentioning

confidence: 99%

Local search five‐element cycle optimized reLU‐BiLSTM for multilingual aspect‐based text classification

Kumar¹,

Sulochana

2022

Concurrency and Computation

View full text Add to dashboard Cite

show abstract

“…Research works in this category demonstrate methods for prioritizing tweets and extracting useful information, such as rescue requests or informative tweets, to aid in disaster response efforts. Within this systematic literature review at least nine studies were categorized within this classification [29,31,[52][53][54][55][56][57][58]. This study critically analyzed 53 published papers and identified 9 belonging into this category (as shown in Table 8).…”

Section: Tweet Prioritization and Useful Information Extractionmentioning

confidence: 99%

Section: Multilingual and Cross-lingual Disaster Analysismentioning

confidence: 99%

“…These are studies addressing disaster analysis and classification across different languages or in multilingual contexts such as the works demonstrated in [55,64] (Table 10). The detailed summaries for both [55,64] are located in Appendix A, Table A7. Achieved top performance in tweet prioritization and outperformed median performance for information type classification using pre-trained language models.…”

Section: Multilingual and Cross-lingual Disaster Analysismentioning

confidence: 99%

See 1 more Smart Citation

A Sustainable Way Forward: Systematic Review of Transformer Technology in Social-Media-Based Disaster Analytics

Sufi

2024

Sustainability

View full text Add to dashboard Cite

Transformer technologies, like generative pre-trained transformers (GPTs) and bidirectional encoder representations from transformers (BERT) are increasingly utilized for understanding diverse social media content. Despite their popularity, there is a notable absence of a systematic literature review on their application in disaster analytics. This study investigates the utilization of transformer-based technology in analyzing social media data for disaster and emergency crisis events. Leveraging a systematic review methodology, 114 related works were collated from popular databases like Web of Science and Scopus. After deduplication and following the exclusion criteria, 53 scholarly articles were analyzed, revealing insights into the geographical distribution of research efforts, trends in publication output over time, publication venues, primary research domains, and prevalently used technology. The results show a significant increase in publications since 2020, with a predominant focus on computer science, followed by engineering and decision sciences. The results emphasize that within the realm of social-media-based disaster analytics, BERT was utilized in 29 papers, BERT-based methods were employed in 28 papers, and GPT-based approaches were featured in 4 papers, indicating their predominant usage in the field. Additionally, this study presents a novel classification scheme consisting of 10 distinct categories that thoroughly categorize all existing scholarly works on disaster monitoring. However, the study acknowledges limitations related to sycophantic behavior and hallucinations in GPT-based systems and raises ethical considerations and privacy concerns associated with the use of social media data. To address these issues, it proposes strategies for enhancing model robustness, refining data validation techniques, and integrating human oversight mechanisms.

show abstract

“…GNNs can be trained to identify key terms and concepts within documents across multiple languages and map them to their corresponding translations in English. This facilitates the extraction of important information from multilingual sources and supports the generation of high-quality translations that capture the essence of the original content [11].Additionally, GNNs can be applied to cross-lingual sentence embedding, where sentences from different languages are embedded into a shared semantic space [12]. This enables more effective comparison and matching of sentences across languages, facilitating tasks such as cross-lingual paraphrase identification and translation alignment, which are essential for accurate English translation [13].The integration of graph neural networks into multilingual information retrieval systems holds great promise for advancing the field of English translation.…”

Section: Introductionmentioning

confidence: 99%

Multilingual Information Retrieval Using Graph Neural Networks: Practical Applications in English Translation

Hao Yang

2024

jes

View full text Add to dashboard Cite

Multilingual information retrieval using graph neural networks offers practical applications in English translation by leveraging advanced computational models to enhance the efficiency and accuracy of cross-lingual search and translation tasks. By representing textual data as graphs and utilizing graph neural networks (GNNs), this approach captures intricate relationships between words and phrases across different languages, enabling more effective language understanding and translation. GNNs can learn complex linguistic structures and semantic similarities from multilingual corpora, facilitating the development of more robust translation systems that are capable of handling diverse language pairs and domains. The paper introduces a novel approach termed the Multilingual Ant Bee Optimization Graph Neural Network (MABO-GNN) for addressing optimization, classification, and multilingual translation tasks. MABO-GNN integrates ant bee optimization algorithms with graph neural networks to provide a versatile framework capable of optimizing objective functions, improving classification accuracy iteratively, and facilitating high-quality translations across multiple languages. Through comprehensive experimentation, the efficacy of MABO-GNN is demonstrated across various tasks, languages, and datasets. in optimization experiments, MABO-GNN achieves objective function values of 0.012, 0.015, 0.011, and 0.013 in Experiment 1, Experiment 2, Experiment 3, and Experiment 4, respectively, with convergence times ranging from 90 to 150 seconds. In classification tasks, the model exhibits notable performance improvements over iterations, with BLEU scores reaching 0.84 and METEOR scores reaching 0.78 in the fifth iteration. The translation results showcase BLEU scores of 0.85 for English, 0.82 for French, 0.79 for German, 0.81 for Spanish, and 0.75 for Chinese, indicating the model's proficiency in generating high-quality translations across diverse languages.

show abstract

GNoM: Graph Neural Network Enhanced Language Models for Disaster Related Multilingual Text Classification

Cited by 12 publications

References 13 publications

Local search five‐element cycle optimized reLU‐BiLSTM for multilingual aspect‐based text classification

Local search five‐element cycle optimized reLU‐BiLSTM for multilingual aspect‐based text classification

A Sustainable Way Forward: Systematic Review of Transformer Technology in Social-Media-Based Disaster Analytics

Multilingual Information Retrieval Using Graph Neural Networks: Practical Applications in English Translation

Contact Info

Product

Resources

About