Search Results (64)

Text summarization is a complex and essential task in natural language processing (NLP) research, focused on extracting the most important information from a document. This study focuses on the Extractive and Abstractive approaches of Bengali Text Summarization (BTS). With the breakthrough advancements in deep learning, summarization is no longer a major challenge for English, given the availability of extensive resources dedicated to this global language. However, the Bengali language remains underexplored. Hence, in this work, a comprehensive review has been conducted on BTS research from 2007 to 2023, analyzing trends, datasets, preprocessing techniques, methodologies, evaluations, and challenges. Leveraging 106 journal and conference papers, this review offers insights into emerging topics and trends in Bengali Abstractive summarization. The review has been augmented with experiments using transformer models from Hugging Face and publicly available datasets to assess the Rouge score accuracy for Abstractive summarization. The extensive literature review conducted in this study reveals that before the advent of transformers, LSTM (Long Short-Term Memory) models were the dominant deep learning approach for text summarization across various languages. For transformers, one of the key datasets utilized was XL-SUM with the MT5 model emerging as the best performer among various contemporary multilingual models. These findings contribute to understanding the contemporary techniques and challenges in BTS. Furthermore, recommendations are made to guide future research endeavors, aiming to provide valuable insights and directions for researchers in this field. Full article

Emerging from the Bengali Muslim char-dwellers in the riverine environments of the Brahmaputra and its tributaries, the Miya Poetry movement is a unique environmentalism of the marginalized in contemporary Assam, India. Writing as a native scholar of Assam, I look at how the poetry movement displays the ethos of an ecopolitical spirituality that embodies the riverine ecology, environmental politics, and sacrality and how it challenges the majoritarian state’s narrative of the Bengali Muslim char-dwellers being denigrated as the “environmental waste producers”. My concept of “ecopolitical spirituality” is in tandem with Carol White’s ‘African American religious naturalism’, which elucidates the remembrance and evocation of traditional environmental relationships of and by the marginalized communities with the purpose of healing and rehumanizing themselves. I begin with a short history of the Miya Poetry movement among the Bengali Muslim char-dwellers in Assam. It narrates how the leading Miya poets adopt the local “Miya” dialect to express the traditional and continued relationships of Bengali Muslim char-dwellers who find themselves entangled with and nurtured by the land, rivers, plants, and animals. I then examine how Bengali Muslims have been framed by the majoritarian state and Assamese society as “environmental waste producers”. With climate change-induced destructive floods, along with post-colonial state’s rampant building of embankments leading to violent floods and erosion, Bengali Muslim char-dwellers are forced to migrate to nearby government grazing reserves or national parks. There, the majoritarian state projects them to be damaging the environment and issues violent evictions. In state reports too, the Bengali Muslim char-dwellers have been equated with “rats”, “crows”, and “vultures”. I use the concept of “environmental racism” to show how this state-led denigration justifies the allegation of the Muslim char-dwellers as “environmental waste producers” and how the Miya Poetry movement counters the racist allegation with new metaphors by highlighting the traditional relationships of the marginalized community with the riverine environment. In the final section, I look in detail at the characteristics and reasons that make the poetry movement ecopolitically spiritual in nature. I thus lay out an argument that the ecopolitical spirituality of the Miya Poetry movement resists the statist dehumanization and devaluation of Miya Muslims by not mocking, violating, or degrading the majoritarian Assamese but by rehumanizing themselves and their relationship with the environment. Full article

Social media has increasingly relied on memes as a tool for expressing opinions, making meme sentiment analysis an emerging area of interest for researchers. While much of the research has focused on English-language memes, under-resourced languages, such as Bengali, have received limited attention. Given the surge in social media use, the need for sentiment analysis of memes in these languages has become critical. One of the primary challenges in this field is the lack of benchmark datasets, particularly in languages with fewer resources. To address this, we used the MemoSen dataset, designed for Bengali, which consists of 4368 memes annotated with three sentiment labels: positive, negative, and neutral. MemoSen is divided into training (70%), test (20%), and validation (10%) sets, with an imbalanced class distribution: 1349 memes in the positive class, 2728 in the negative class, and 291 in the neutral class. Our approach leverages advanced deep learning techniques for multimodal sentiment analysis in Bengali, introducing three hybrid approaches. SentimentTextFormer is a text-based, fine-tuned model that utilizes state-of-the-art transformer architectures to accurately extract sentiment-related insights from Bengali text, capturing nuanced linguistic features. SentimentImageFormer is an image-based model that employs cutting-edge transformer-based techniques for precise sentiment classification through visual data. Lastly, SentimentFormer is a hybrid model that seamlessly integrates both text and image modalities using fusion strategies. Early fusion combines textual and visual features at the input level, enabling the model to jointly learn from both modalities. Late fusion merges the outputs of separate text and image models, preserving their individual strengths for the final prediction. Intermediate fusion integrates textual and visual features at intermediate layers, refining their interactions during processing. These fusion strategies combine the strengths of both textual and visual data, enhancing sentiment analysis by exploiting complementary information from multiple sources. The performance of our models was evaluated using various accuracy metrics, with SentimentTextFormer achieving 73.31% accuracy and SentimentImageFormer attaining 64.72%. The hybrid model, SentimentFormer (SwiftFormer with mBERT), employing intermediate fusion, shows a notable improvement in accuracy, achieving 79.04%, outperforming SentimentTextFormer by 5.73% and SentimentImageFormer by 14.32%. Among the fusion strategies, SentimentFormer (SwiftFormer with mBERT) achieved the highest accuracy of 79.04%, highlighting the effectiveness of our fusion technique and the reliability of our multimodal framework in improving sentiment analysis accuracy across diverse modalities. Full article

This case study examined the process of integration of cognitive behavioral therapy (CBT) and sensorimotor psychotherapy (SP) in three-phase trauma treatment with a 32-year-old female Bengali refugee with Complex Post Traumatic Disorder (C-PTSD). The treatment was provided in a Western country. The client’s improvement was assessed by using self-report assessments of post-traumatic, dissociative, and depressive symptoms. Psychodiagnostics’ scores gathered after 2 years of treatment showed significant improvements in post-traumatic, dissociative, and depressive symptoms. Although firm conclusions cannot be drawn due to the limitations of this study, results suggest that integrating CBT and SP in a phase-trauma approach may be an effective treatment for C-PTSD in a transcultural setting. However, replicating and standardizing such preliminary results on larger samples is necessary. Background/Objectives: SP is an innovative psychotherapeutic intervention aimed at treating trauma through a bottom-up approach, however, little research exists regarding the efficacy of this psychotherapy. This case study aims to investigate the efficacy of the combination of CBT and SP in treating a C-PTSD patient in a transcultural setting. Methods: Three questionnaires were administered to investigate post-traumatic, depressive, and dissociative symptoms. Results: Clinical and psychodiagnostic outcomes highlight significant psychological improvements in the patient. Conclusions: Although any firm conclusion cannot be drawn because of various intrinsic limitations (i.e., single case study) that straiten our ability to extend these results, this case study suggests that the integration of CBT and SP may be an effective treatment for C-PTSD in a transcultural setting. Full article

This article locates the clinic as a historically contingent space which faced cultural resistance and remained alien to the colonized population in India. It corroborates the socio-political tension in setting up a clinic within the colony and investigates how the Western clinic as a colonial apparatus was resituated as the “Indian clinic” per se. With the historical emergence of a new class of medical practitioners called “daktars” (a Bengali vernacularization of the term “doctor”), the health-seeking behaviour and public health model of colonial India witnessed a decolonial shift. Unlike their English counterparts, daktars did not enjoy a privileged position within the medical archives of colonial India. This archival gap within Indian medical history presents itself as a viable topic for discussion through the means of the literature of the colonized. Bengali writer Ashoke Mukhopadhyay’s novel Abiram Jwarer Roopkotha (2018), translated into English as A Ballad of Remittent Fever in 2020, remedies the colonial politics of the archive by reconstructing the lives of various daktars and their pursuit of self-reliance. The article takes a neo-historical approach towards understanding and assessing the past of daktari medicine and thereby offers comments on its traces in the contemporary public health of India. Full article

The rise in abusive language on social media is a significant threat to mental health and social cohesion. For Bengali speakers, the need for effective detection is critical. However, current methods fall short in addressing the massive volume of content. Improved techniques are urgently needed to combat online hate speech in Bengali. Traditional machine learning techniques, while useful, often require large, linguistically diverse datasets to train models effectively. This paper addresses the urgent need for improved hate speech detection methods in Bengali, aiming to fill the existing research gap. Contextual understanding is crucial in differentiating between harmful speech and benign expressions. Large language models (LLMs) have shown state-of-the-art performance in various natural language tasks due to their extensive training on vast amounts of data. We explore the application of LLMs, specifically GPT-3.5 Turbo and Gemini 1.5 Pro, for Bengali hate speech detection using Zero-Shot and Few-Shot Learning approaches. Unlike conventional methods, Zero-Shot Learning identifies hate speech without task-specific training data, making it highly adaptable to new datasets and languages. Few-Shot Learning, on the other hand, requires minimal labeled examples, allowing for efficient model training with limited resources. Our experimental results show that LLMs outperform traditional approaches. In this study, we evaluate GPT-3.5 Turbo and Gemini 1.5 Pro on multiple datasets. To further enhance our study, we consider the distribution of comments in different datasets and the challenge of class imbalance, which can affect model performance. The BD-SHS dataset consists of 35,197 comments in the training set, 7542 in the validation set, and 7542 in the test set. The Bengali Hate Speech Dataset v1.0 and v2.0 include comments distributed across various hate categories: personal hate (629), political hate (1771), religious hate (502), geopolitical hate (1179), and gender abusive hate (316). The Bengali Hate Dataset comprises 7500 non-hate and 7500 hate comments. GPT-3.5 Turbo achieved impressive results with 97.33%, 98.42%, and 98.53% accuracy. In contrast, Gemini 1.5 Pro showed lower performance across all datasets. Specifically, GPT-3.5 Turbo excelled with significantly higher accuracy compared to Gemini 1.5 Pro. These outcomes highlight a 6.28% increase in accuracy compared to traditional methods, which achieved 92.25%. Our research contributes to the growing body of literature on LLM applications in natural language processing, particularly in the context of low-resource languages. Full article

Sign language is undoubtedly a common way of communication among deaf and non-verbal people. But it is not common among hearing people to use sign language to express feelings or share information in everyday life. Therefore, a significant communication gap exists between deaf and hearing individuals, despite both groups experiencing similar emotions and sentiments. In this paper, we developed a convolutional neural network–squeeze excitation network to predict the sign language signs and developed a smartphone application to provide access to the ML model to use it. The SE block provides attention to the channel of the image, thus improving the performance of the model. On the other hand, the smartphone application brings the ML model close to people so that everyone can benefit from it. In addition, we used the Shapley additive explanation to interpret the black box nature of the ML model and understand the models working from within. Using our ML model, we achieved an accuracy of 99.86% on the KU-BdSL dataset. The SHAP analysis shows that the model primarily relies on hand-related visual cues to predict sign language signs, aligning with human communication patterns. Full article

An English-Bengali machine translation (MT) application can convert an English text into a corresponding Bengali translation. To build a better model for this task, we can optimize English-Bengali MT. MT for languages with rich resources, like English-German, started decades ago. However, MT for languages lacking many parallel corpora remains challenging. In our study, we employed back-translation to improve the translation accuracy. With back-translation, we can have a pseudo-parallel corpus, and the generated (pseudo) corpus can be added to the original dataset to obtain an augmented dataset. However, the new data can be regarded as noisy data because they are generated by models that may not be trained very well or not evaluated well, like human translators. Since the original output of a translation model is a probability distribution of candidate words, to make the model more robust, different decoding methods are used, such as beam search, top-k random sampling and random sampling with temperature T, and others. Notably, top-k random sampling and random sampling with temperature T are more commonly used and more optimal decoding methods than the beam search. To this end, our study compares LSTM (Long-Short Term Memory, as a baseline) and Transformer. Our results show that Transformer (BLEU: $27.80$ in validation, $1.33$ in test) outperforms LSTM ($3.62$ in validation, $0.00$ in test) by a large margin in the English-Bengali translation task. (Evaluating LSTM and Transformer without any augmented data is our baseline study.) We also incorporate two decoding methods, top-k random sampling and random sampling with temperature T, for back-translation that help improve the translation accuracy of the model. The results show that data generated by back-translation without top-k or temperature sampling (“no strategy”) help improve the accuracy (BLEU $38.22$, +$10.42$ on validation, $2.07$, +$0.74$ on test). Specifically, back-translation with top-k sampling is less effective ($k=10$, BLEU $29.43$, +$1.83$ on validation, $1.36$, +$0.03$ on test), while sampling with a proper value of T, $T=0.5$ makes the model achieve a higher score ($T=0.5$, BLEU $35.02$, +$7.22$ on validation, $2.35$, +$1.02$ on test). This implies that in English-Bengali MT, we can augment the training set through back-translation using random sampling with a proper temperature T. Full article

The cause(s) of missing inflectional morphology in obligatory contexts by adult speakers of second language (L2) English is subject to ongoing discussion. Whatever the specific theory, however, the apparent asymmetrical production of the morpheme ‘-s’ in the marking of number on plural nouns versus that on third person singular agreement has to be accounted for. This study adopts the theoretical approach put forward by the Prosodic Transfer Hypothesis, whereby the prosodic representation of inflectional morphology in the first language (L1) can, to some extent, account for differences in the suppliance of inflectional morphology in L2 English within and across L1s. It is in this context that the production of third person singular agreement, and, for contrastive purposes, number on plural nouns, by L1 Bengali speakers of L2 English, is considered in relation to available prosodic representation in the L1, as well as against phonological processes attested in L1 acquisition. More specifically, with respect to covert contrasts. An inspection of spectrograms from instances of the apparent omission of inflection by L1 Bengali speakers of L2 English at Beginner to Intermediate proficiency levels does not, however, indicate that learners are covertly supplying agreement on the third person singular (or plural number on nouns). This finding does not necessarily rule out the occurrence of covert contrasts in L2 production of inflectional morphology; alternative techniques may detect a systematic difference between bare verbs and non-audible (to the listener) inflection. Full article

Undertaken in Kolkata, India, our study aimed to explore the experiences of Bengali middle-class women on perceived stressful events, social support, and coping experiences following childbirth. Becoming a mother following childbirth is a shared phenomenon irrespective of culture, social strata, or country, while stress during the postpartum period or depression is not. Discrete medical intervention does not sufficiently address the complexities of postpartum experiences since influencing factors also include economic, political, cultural, and social backgrounds. Adopting a feminist and phenomenological approach, individual in-person interviews were conducted with twenty women recruited via snowball sampling. Our findings revealed that events experienced as stressful may lead to poor postpartum well-being. Underpinned by gendered discourse and biases, stressful events included familial imperatives for a male child, poor social and emotional support from the family, mostly partners and fathers, and systemic workplace barriers. The women in our study commonly resided with their mothers postpartum. They expressed feeling sheltered from these experiences, cared for, and supported. We discuss the women’s experiences from a feminist pragmatic worldview, which advocates for a flexible feminism recognizant of the unique and nurturing relationship experiences between Bengali middle-class women and their mothers. In conclusion, we advocate for culturally sensitive, women-centered postpartum care practices that may entail the inclusion of intergenerational care during this critical phase of maternal well-being. These insights underscore the necessity of tailoring postpartum support systems to align with the cultural and familial contexts of the individuals they serve. Full article

This study aimed to identify the impact of English language proficiency on Bangladeshi graduates in terms of its influence on local languages and cultural integration. The study was conducted using a quantitative approach, and the random sampling technique was employed to select 370 respondents from English and Bengali medium-instructed Bangladeshi graduates. Data were analysed using SPSS version 27.0. The Pearson correlation, ANOVA, and multiple regression analyses were used in line with the research objectives. The Pearson correlation shows a significant correlation among the studied variables, such as cultural erosion (R² = 047, p < 0.001), threat to local language (R² = 048, p < 0.001), and polluting local language (R² = 047, p < 0.001). The ANOVA test was performed to find the mean difference in cultural erosion (CE), threat to local language (TLL), and polluting local language (PLL) with respect to respondents’ qualifications (bachelor’s, master’s, or Ph.D. degree). The ANOVA result revealed no statistically significant difference in CE and TLL in terms of graduates’ qualifications, while PLL differs significantly based on their qualifications (bachelor’s, master’s, or Ph.D. degree.). Similarly, significant differences in CE, TLL, and PLL were found among the income groups and living places of graduates. Multiple regression analysis explained 13.6% variances in the social class context with ELP-linked challenges (R² = 0.136, p = 0.001), while the lower class was found to have positive non-significant relationships with ELP-linked challenges (B = 0. 159, p = 0.382). This was almost double that of the result for the rich class (B = 0.085, p = 0.721). However, ELP-linked challenges contain a mixture of languages and cultural blending in society, as the rich class faces fewer challenges than the lower and middle classes. The study is influential in creating an awareness of language use whenever necessary. Full article

Saradamani Mukhopadhyay (1853–1920), more widely known as Sarada Devi and, to her devotees, the Holy Mother, presents an illuminating case study of the various means by which, in many respects, a highly traditional and typical rural Hindu woman of her time, operating from within the categories of Bengali Hindu society, was able to navigate these categories in ways that did not undermine, but rather enhanced, her agency, enabling her to shape her social reality in creative and transformative ways. Inhabiting the traditional role of mother and nurturer while carrying it out in a highly non-traditional manner, Sarada Devi played a central, often behind-the-scenes, role as a major influencer of an important modern Hindu spiritual movement—indeed, the first such movement to be able to boast an international following. Having no biological children of her own, Sarada Devi became the mother to this movement and to the monastic order dedicated to carrying forward the vision of her husband, Sri Ramakrishna Paramahamsa (1836–1886), as interpreted both by herself and his disciples, the most prominent of whom was Swami Vivekananda (1863–1902), who is well known for having brought Ramakrishna’s teachings to the Western world through his lectures in America, including at the World Parliament of Religions held in Chicago in 1893 and his founding of the first Vedanta Societies, starting in New York in 1894. Full article

Wound therapies involving gene delivery to the skin have significant potential due to the advantage and ease of local treatment. However, choosing the appropriate vector to enable successful gene expression while also ensuring that the treatment’s immediate material components are conducive to healing itself is critical. In this study, we utilized a particulate formulation of the polymer chitosan (chitosan particles, CPs) as a non-viral vector for the delivery of a plasmid encoding human CA5-HIF-1α, a degradation resistant form of HIF-1α, to enhance wound healing. We also compared the angiogenic potential of our treatment (HIF/CPs) to that of chitosan particles containing only the plasmid backbone (bb/CPs) and the chitosan particle vector alone (CPs). Our results indicate that chitosan particles exert angiogenic effects that are enhanced with the human CA5-HIF-1α-encoded plasmid. Moreover, HIF/CPs enhanced wound healing in diabetic db/db mice (p < 0.01), and healed tissue was found to contain a significantly increased number of blood vessels compared to bb/CPs (p < 0.01), CPs (p < 0.05) and no-treatment groups (p < 0.01). Thus, this study represents a method of gene delivery to the skin that utilizes an inherently pro-wound-healing polymer as a vector for plasmid DNA that has broad application for the expression of other therapeutic genes. Full article

Fish has always been an integral part of Bengali cuisine and economy. Fish could also be a potential reservoir of pathogens. This study aimed to inquisite the distribution of virulence, biofilm formation, and antimicrobial resistance of Enterococcus faecalis isolated from wild and cultivated fish in Bangladesh. A total of 132 koi fish (Anabas scandens) and catfish (Heteropneustes fossilis) were collected from different markets in the Mymensingh district and analyzed to detect E. faecalis. E. faecalis was detected by conventional culture and polymerase chain reaction (PCR), followed by the detection of virulence genes by PCR. Antibiotic susceptibility was determined using the disk diffusion method, and biofilm-forming ability was investigated by crystal violet microtiter plate (CVMP) methods. A total of 47 wild and 40 cultured fish samples were confirmed positive for E. faecalis by PCR. The CVMP method revealed four per cent of isolates from cultured fish as strong biofilm formers, but no strong producers were found from the wild fish. In the PCR test, 45% of the isolates from the wild and cultivated fish samples were found to be positive for at least one biofilm-producing virulence gene, where agg, ace, gelE, pil, and fsrC genes were detected in 80, 95, 100, 93, and 100% of the isolates, respectively. Many of the isolates from both types of samples were multidrug resistant (MDR) (73% in local fish and 100% in cultured fish), with 100% resistance to erythromycin, linezolid, penicillin, and rifampicin in E. faecalis from cultured fish and 73.08, 69.23, 69.23, and 76.92%, respectively, in E. faecalis from wild fish. This study shows that E. faecalis from wild fish have a higher frequency of virulence genes and biofilm-forming genes than cultivated fish. However, compared to wild fish, cultured fish were found to carry E. faecalis that was more highly multidrug resistant. Present findings suggest that both wild and cultured fish could be potential sources for MDR E. faecalis, having potential public health implications. Full article

The primary goal of Liberation Theology is to change the material conditions of marginalised and oppressed groups in society. Within Islamic Liberation Theology, however, issues related to class and economic inequality are notably missing. This paper seeks to begin this conversation and highlight the necessity of addressing economic exploitation, which affects most of the world’s population and Muslims disproportionately. Using a praxis-based methodology, it centres the interpretation of activists from Nijjor Manush, a British Bengali activist group, and seeks to understand how Islam is used as a liberative tool to combat class oppression. Through interviews and focus groups, an alternative and revolutionary Islam emerges. Echoing a Marxist understanding of class, it sees exploitation as an inherent part of the current capitalist system and recognises the necessity of people seizing economic power. This overarching objective is the lens through which activism in the here and now is interpreted and tactics decided. Establishing economic justice therefore means trying to secure “non-reformist reforms” in the short term, which resist the logic of capital and secure the interests of the marginalised, while working towards the ultimate goal of ending economic exploitation and, by extension, abolishing class. Full article