Model Of Language Research Articles

EHMMQA: English, Hindi, and Marathi multilingual question answering framework using deep learning

Abstract Multilingual question answering (MQA) is an effective access to multilingual data to provide accurate and precise answers, irrespective of language. Although a wide range of datasets is available for monolingual QA systems in natural language processing, benchmark datasets specifically designed for MQA are considerably limited. The absence of comprehensive and benchmark datasets hinders the development and evaluation of MQA systems. To overcome this issue, the proposed work attempts to develop the EHMQuAD dataset, an MQA dataset for low-resource languages such as Hindi and Marathi accompanying the English language. The EHMQuAD dataset is developed using a synthetic corpora generation approach, and an alignment is performed after translation to make the dataset more accurate. Further, the EHMMQA model is proposed to create an abstract framework that uses a deep neural network that accepts pairs of questions and context and returns an accurate answer based on those questions. The shared question and shared context representation have been designed separately to develop this system. The experiments of the proposed model are conducted on the MMQA, Translated SQuAD, XQuAD, MLQA, and EHMQuAD datasets, and EM and F1-score are used as performance measures. The proposed model (EHMMQA) is collated with state-of-the-art MQA baseline models for all possible monolingual and multilingual settings. The results signify that EHMMQA is a considerable step toward the MQA system utilizing Hindi and Marathi languages. Hence, it becomes a new state-of-the-art model for Hindi and Marathi languages.

Salvation, Enlightenment and Awakening in Shaping Followers' Perception and Understanding of Their Faith: A Linguistic Exploration

This paper examines how religious concepts of salvation, enlightenment, and awakening shape followers’ perception and understanding of their faith. Drawing on theories of language ideology (Woolard, 1998) and linguistic structuration (Giddens, 1979), it explores how linguistic and conceptual variations surrounding these key terms influence religious worldviews. A mixed methods approach integrates comparative-historical case studies, interviews, surveys, and textual analysis of three religious’ denominations. Findings reveal ideological divergences in the metaphorical framing of the key concepts, relating to differences in perceived resonances and definitions among followers. The study demonstrates that religious conceptual systems transmit between contexts via semiotic processes of indexicality and erasure, undergoing shifts in meaning. These ideological variations in turn contribute to divergent understandings of core principles across faith traditions. Ultimately, illuminating conceptual divergences and the symbolic processes by which they spread provides vital insight into religion’s cultural logic and diversity. The research contributes to theoretical models of religious language’s co-constitutive relationship with thought and experience. It offers an empirical approach employing cognitive linguistics and the sociology of religion to promote understanding across belief systems in an increasingly pluralistic world.

Preservation of sentiment in machine translation of low-resource languages: a case study on Slovak movie subtitles

AbstractThis research explores the effectiveness of machine translation from Slovak to English for sentiment analysis, specifically focusing on the translation of movie subtitles. The study employs a parallel corpus of segmented movie subtitles in both languages and utilizes IBM Watson™ Natural Language Understanding service and Google Translate. The research aims to assess the correlation between human-generated text and machine-translated text in sentiment analysis. A comparative analysis was also conducted using OpenAI to evaluate the sentiment of the Slovak text directly, without translation into English. The findings reveal a strong correlation between human text and machine translation, with a Pearson correlation coefficient of 0.86, and a correlation with OpenAI’s GPT model evaluation at 0.72. Despite the relatively high accuracy of the end-to-end solution using OpenAI, the methodology comprising machine translation followed by sentiment analysis in English was found to be significantly more precise. The research further investigates the challenges in translating specific language nuances, such as humor and vulgarism, and their impact on sentiment analysis. The study concludes that machine translation can be effectively used for sentiment analysis in Slovak, a flective language, and highlights the potential of advanced language models in low-resource languages. Future research directions include expanding the study to other text types and comparable languages beyond Slovak.

Semantically Enriched Cross-Lingual Sentence Embeddings for Crisis-related Social Media Texts

Tasks such as semantic search and clustering on crisis-related social media texts enhance our comprehension of crisis discourse, aiding decision-making and targeted interventions. Pre-trained language models have advanced performance in crisis informatics, but their contextual embeddings lack semantic meaningfulness. Although the CrisisTransformers family includes a sentence encoder to address the semanticity issue, it remains monolingual, processing only English texts. Furthermore, employing separate models for different languages leads to embeddings in distinct vector spaces, introducing challenges when comparing semantic similarities between multi-lingual texts. Therefore, we propose multi-lingual sentence encoders (CT-XLMR-SE and CT-mBERT-SE) that embed crisis-related social media texts for over 50 languages, such that texts with similar meanings are in close proximity within the same vector space, irrespective of language diversity. Results in sentence encoding and sentence matching tasks are promising, suggesting these models could serve as robust baselines when embedding multi-lingual crisis-related social media texts. The models are publicly available at: https://huggingface.co/crisistransformers.

Multilingual Neural Machine Translation for Indic to Indic Languages

The method of translation from one language to another without human intervention is known as Machine Translation (MT). Multilingual neural machine translation (MNMT) is a technique for MT that builds a single model for multiple languages. It is preferred over other approaches, since it decreases training time and improves translation in low-resource contexts, i.e., for languages that have insufficient corpus. However, good-quality MT models are yet to be built for many scenarios such as for Indic-to-Indic Languages (IL-IL). Hence, this article is an attempt to address and develop the baseline models for low-resource languages i.e., IL-IL (for 11 Indic Languages (ILs)) in a multilingual environment. The models are built on the Samanantar corpus and analyzed on the Flores-200 corpus. All the models are evaluated using standard evaluation metrics i.e., Bilingual Evaluation Understudy (BLEU) score (with the range of 0 to 100). This article examines the effect of the grouping of related languages, namely, East Indo-Aryan (EI), Dravidian (DR), and West Indo-Aryan (WI) on the MNMT model. From the experiments, the results reveal that related language grouping is beneficial for the WI group only while it is detrimental for the EI group and it shows an inconclusive effect on the DR group. The role of pivot-based MNMT models in enhancing translation quality is also investigated in this article. Owing to the presence of large good-quality corpora from English (EN) to ILs, MNMT IL-IL models using EN as a pivot are built and examined. To achieve this, English-Indic Language (EN-IL) models are developed with and without the usage of related languages. Results show that the use of related language grouping is advantageous specifically for EN to ILs. Thus, related language groups are used for the development of pivot MNMT models. It is also observed that the usage of pivot models greatly improves MNMT baselines. Furthermore, the effect of transliteration on ILs is also analyzed in this article. To explore transliteration, the best MNMT models from the previous approaches (in most of cases pivot model using related groups) are determined and built on corpus transliterated from the corresponding scripts to a modified Indian language Transliteration script (ITRANS). The outcome of the experiments indicates that transliteration helps the models built for lexically rich languages, with the best increment of BLEU scores observed in Malayalam (ML) and Tamil (TA), i.e., 6.74 and 4.72, respectively. The BLEU score using transliteration models ranges from 7.03 to 24.29. The best model obtained is the Punjabi (PA)-Hindi (HI) language pair trained on PA-WI transliterated corpus.

The Role of Artificial Intelligence in English Language and Literature Reading Management

Firstly, this paper analyzes the role of AI in the reading management of English language and literature, establishes the implicit knowledge base of neural network, designs the auxiliary reading system for learning English language and literature, and optimizes the English language and literature management model of AI. The experimental results show that its reading efficiency is increased by 0.48%, and the performance of the credibility model is improved by 0.53% compared with the original system, which greatly optimizes the running time of the system. To some extent, it helps users to manage their time in English language and literature reading, and greatly improves users' reading efficiency and quality. Based on this advantage of AI algorithm, this paper introduces that the algorithm optimizes the reading management model and the training process of neural grid, and constructs a model of English language and literature assisted reading system based on AI. The system can better meet the needs of users in English language and literature reading management.

Amharic spoken digits recognition using convolutional neural network

Spoken digits recognition (SDR) is a type of supervised automatic speech recognition, which is required in various human–machine interaction applications. It is utilized in phone-based services like dialing systems, certain bank operations, airline reservation systems, and price extraction. However, the design of SDR is a challenging task that requires the development of labeled audio data, the proper choice of feature extraction method, and the development of the best performing model. Even if several works have been done for various languages, such as English, Arabic, Urdu, etc., there is no developed Amharic spoken digits dataset (AmSDD) to build Amharic spoken digits recognition (AmSDR) model for the Amharic language, which is the official working language of the government of Ethiopia. Therefore, in this study, we developed a new AmSDD that contains 12,000 utterances of 0 (Zaero) to 9 (zet’enyi) digits which were recorded from 120 volunteer speakers of different age groups, genders, and dialects who repeated each digit ten times. Mel frequency cepstral coefficients (MFCCs) and Mel-Spectrogram feature extraction methods were used to extract trainable features from the speech signal. We conducted different experiments on the development of the AmSDR model using the AmSDD and classical supervised learning algorithms such as Linear Discriminant Analysis (LDA), K-Nearest Neighbors (KNN), Support Vector Machine (SVM), and Random Forest (RF) as the baseline. To further improve the performance recognition of AmSDR, we propose a three layers Convolutional Neural Network (CNN) architecture with Batch normalization. The results of our experiments show that the proposed CNN model outperforms the baseline algorithms and scores an accuracy of 99% and 98% using MFCCs and Mel-Spectrogram features, respectively.

Enhancing abusive language detection: A domain-adapted approach leveraging BERT pre-training tasks

The widespread adoption of deep learning approaches in natural language processing is largely attributed to their exceptional performance across diverse tasks. Notably, Transformer-based models, such as BERT, have gained popularity for their remarkable efficacy and their ease of adaptation (via fine-tuning) across various domains. Despite their success, fine-tuning these models for informal language, particularly instances involving offensive expressions, presents a major challenge due to limitations in vocabulary coverage and contextual information for such tasks. To address these challenges, we propose the domain adaptation of the BERT language model for the task of detecting abusive language. Our approach involves constraining the language model with the adaptation and paradigm shift of two default pre-trained tasks, the design of two datasets specifically engineered to support the adapted pre-training tasks, and the proposal of a dynamic weighting loss function. The evaluation of these adapted configurations on six datasets dedicated to abusive language detection reveals promising outcomes, with a significant enhancement observed compared to the base model. Furthermore, our proposed methods yield competitive results when compared to state-of-the-art approaches, establishing a robust and easily trainable model for the effective identification of abusive language.

Synchronous integration method of mechatronic system design, geometric design, and simulation based on SysML

The benefits of integrated design using the Model Based System Engineering (MBSE) approach in the design process of mechatronic systems have gradually become apparent. The automatic generation of simulation models and geometric models based on System Modeling Languages (SysML) models enables system engineers to swiftly analyze and simulate system performance, visually depict design outcomes, and expedite the product development process. Due to the current system modeling's lack of model integration and geometric design functions, this paper proposes an integrated design and simulation method for mechatronic systems that can carry out complete model synchronization and verification, rapid geometric solution generation, and visual representation. Furthermore, a corresponding model synchronization integration framework is established. This framework primarily encompasses system de-sign, system simulation, and geometric design, with its model integration method being model transformation and model synchronization. The paper concludes with an example of the design process of a quadruped robot to validate the framework and its supported methods, providing a reference for other system design and integration endeavors.

The model of tandem content and language integrated learning: Justification and evaluation of its effectiveness by education stakeholders

Introduction. The need to master a foreign language in the professional field of education is determined by contemporary challenges: the rapid growth of intercultural interaction, as well as the processes of globalization and informatization of education. The article examines the problem of using tandem content and language integrated learning focused on developing future teachers’ readiness and ability to communicate effectively in a foreign language in the professional environment. The purpose of the article is to propose the model of tandem content and language integrated learning and evaluate the effectiveness of its implementation in the process of developing foreign language professional communicative competence of future teachers. Materials and Methods. The study was conducted during 2020–2023 period. The sample included 86 students of 3–5 years majoring in Primary education and English as a Foreign language. The authors used the following research methods: theoretical and methodological analysis, comparison, interpretation, modeling, questionnaires, analysis and systematization of accumulated experience of tandem content and language integrated learning. Results. The main results consist in substantiating the model of tandem content and language integrated learning and in assessing the effectiveness of its implementation in the process of developing foreign language professional communicative competence of future teachers. The authors identified a mechanism of interaction between all education stakeholders in dynamic groups based on task level-differentiation within the framework of content and language integrated learning (CLIL). Conclusions. The study concludes that the model of tandem content and language integrated learning was substantiated and the effectiveness of its implementation in the process of developing foreign language professional communicative competence of future teachers was assessed.

Deep Learning for Arabic Speech Recognition Using Convolutional Neural Networks

Extracting the speaker's emotional state from their speech has become an active research topic lately due to the demand for more human interactive applications. This field of research has noted significant advancement, especially in the English language, owing to the availability of massive speech-labeled corpora. However, the progress of analogous methodologies in the Arabic language is still in its infancy stages. In this paper, we present a Speech Recognition model for the Arabic language, proficient in discerning both the emotional state and gender of the speaker through voice analysis. Three primary emotion labels were selected: low, standard, and high levels of emotion. Various spectral features, such as the mel-frequency cepstral coefficient (MFCC), were extracted and tested to determine the optimal features. Furthermore, various Machine Learning models (SVM, KNN, and HMM) and Deep Learning models (LSTM and CNN) were evaluated for training. The results were compared between the five models using different extracted features, ultimately culminating in the selection of MFCC, root-mean-square (RMS), mel-scaled spectrogram, spectral, and zero-crossing rate as spectral features, and the CNN as a classification model. This selection yielded significant results, with an accuracy of 93% for emotion recognition and 99% for gender recognition.

Neural machine translation for low resource Indian language: Hindi-Kangri

Neural Machine Translation (NMT) for low resource languages is a challenging task due to unavailability of large parallel corpus. The efficacy of Transformer based NMT models largely depends on scale of the parallel corpus and the configuration of hyperparameters implemented during model training. This study aims to delve into and elucidate the impact of hyperparameters on the performance of NMT models for low resource languages. To accomplish this, a series of experiments are conducted using an open-source Hindi-Kangri corpus to train both supervised and semi-supervised NMT models. Throughout the experimentation process, a significant number of discrepancies were identified within the data-set, necessitating manual correction. The best translation performance evaluated with respect to the metrics such as BLEU (0–1), SacreBLEU (0–100), Chrf (0–100), Chrf+ (0–100), Chrf++ (0–100) and TER (%) is (0.15, 14.98, 41.43, 41.49, 38.77, 68.20) for Hindi to Kangri direction, and (0.283, 28.17, 49.71, 50.64, 48.63, 51.25) for Kangri to Hindi direction.

Ensemble random forest and deep convolutional neural networks in detecting and classifying the multiple intrusions from near real‐time cloud datasets

AbstractDue to rapidly growing Internet facilities, intruders can steal and misuse the data saved and stored digitally. In this case, securing digital data is challenging but prominent for various purposes. However, the traditional techniques are insufficient to secure these computer networks and cloud information with a 100% success rate. Recently, machine‐ or deep‐learning‐enabled methods have been used to secure network information, but with some limits. Therefore, the study emphasizes detecting and classifying network intrusion using the proposed ensemble and deep learning models. In this case, we developed the ensemble learning‐enabled random forest algorithm and deep learning‐enabled deep convolutional neural network (CNN) models for securing near real‐time cloud information and designed the intrusion detection system accordingly. The complex and high‐volume CSE‐CICIDS2018 datasets were used to test the developed model in Python programming language implemented with several Python libraries. The outcome of the proposed models indicates that the developed models are promising in securing the cloud information with 97.73% and 99.91% accuracies via ensemble‐random forest and deep CNN models. Thus, the present study models can be applied to other real‐time datasets and computer networks to detect cyber threats effectively.

خصائص اللهجة الكويتية المكتوبة واستخدامها في إنشاء موارد للتحليل الصرفي الآلي.

Kuwaiti Arabic (KA), like other Arabic dialects, is a spoken variety of Arabic that does not have a standardized written convention contrary to Modern Standard Arabic (MSA). With the emergence and spread of social media platforms, Arabic dialects have found their way into the written medium, and hence a need arose to process them alongside MSA. The biggest challenge facing NLP tools is that dialects do not have consistent written conventions contrary to MSA, and writers expressing their dialects usually follow a phonetic writing system, or they write words as they pronounce them. This has opened the door for variations within the same dialect and between dialects and MSA. Furthermore, a prerequisite for analysing any language or dialect is the presence of clear written conventions. Therefore, efforts have been made to establish written conventions for Arabic dialects, but the Kuwaiti dialect has not received the required attention. The current study offers a practical solution for processing written KA. It identified and extracted the written conventions of KA from natural data collected from over 100K Kuwaiti tweets since they represent a good model of natural language. The morphological analyzer (MADAMIRA) - which is designed to process MSA - was enhanced with the extracted criteria. Furthermore, the study involved enriching the analyzer with a dictionary of Kuwaiti terms and vocabulary ‘lemmas’ collected from the Encyclopaedia of Kuwaiti Arabic and from the most used Kuwaiti words on Twitter (currently X). Providing the analyzer with this dictionary of KA words helps it process KA more accurately. The expanded version of the analyzer (MADAMIRA-KA) is the first of its kind designed entirely to process the Kuwaiti dialect and has achieved excellent performance in analyzing over 100K Kuwaiti tweets successfully. The importance of this study lies in developing such a morphological analyzer, which can be used for automated translation, dialect recognition and sentiment analysis.

РЕФАКТОРИНГ КРОС-ПЛАТФОРМНИХ ЗАСТОСУНКІВ З ВИКОРИСТАНЯМ ШТУЧНОГО ІНТЕЛЕКТУ

In software development, the productivity of code refactoring is increased by the use of artificial intelligence (AI). AI methods such as machine learning (ML), natural language processing (NLP), and genetic algorithms (GA) can be used in the refactoring process. Each method has a certain impact on the process, both positive and negative. With this in mind, the work performed by AI requires careful management to avoid a number of problems, such as "hallucinations". Research continues to explore new methods, evaluate comparative effectiveness, and optimize AI models for specific frameworks and languages.

A product language-mediated concept construction model from the perspective of conceptual metaphor

In product design practice, the user experience is directly related to the concepts that users interpret in the product language. Therefore, the designer’s ability to construct concepts through product language is important to the success of the product. This study aimed to establish a conceptual metaphor–based theoretical model that enables designers to construct concepts accurately through product language. Through theory and product analysis, this study explored how designers use product language to generate concepts, specifically through categorization, and explains the core role of conceptual metaphor in the categorization process. A theoretical model of product language–mediated concept construction mechanism was proposed, which was then tested through workshops and semi-structured interviews. Theoretically, the study extends the knowledge of product language to the cognitive level. Practically, the study results will help designers analyze product language and effectively construct and express new concepts through product language.

MULTİKULTURALİZMİN İNKİŞAFINDA DİL AMİLİ

The Linguistic Factor in the Development of Multiculturalism Summary The research conducted has formed an opinion about the importance of the relationship between multiculturalism and language. There is a strong connection between different models of multiculturalism and languages around the world. Because language as a means of communication unites different segments of multiculturalism. If communication is clear and positive, opinions and ideas will be clearly received by different groups and will be useful in negotiations and integration to resolve interests. This, in turn, can create a foundation for peaceful coexistence of people with different identities. In many cases, language problems have been observed in the society. Lack of communication due to language can complicate the problem. Key words: multiculturalism, language, minorities, society, linguistic rights, coexistence

Meta predictive learning model of languages in neural circuits.

Large language models based on self-attention mechanisms have achieved astonishing performances, not only in natural language itself, but also in a variety of tasks of different nature. However, regarding processing language, our human brain may not operate using the same principle. Then, a debate is established on the connection between brain computation and artificial self-supervision adopted in large language models. One of most influential hypotheses in brain computation is the predictive coding framework, which proposes to minimize the prediction error by local learning. However, the role of predictive coding and the associated credit assignment in language processing remains unknown. Here, we propose a mean-field learning model within the predictive coding framework, assuming that the synaptic weight of each connection follows a spike and slab distribution, and only the distribution, rather than specific weights, is trained. This meta predictive learning is successfully validated on classifying handwritten digits where pixels are input to the network in sequence, and moreover, on the toy and real language corpus. Our model reveals that most of the connections become deterministic after learning, while the output connections have a higher level of variability. The performance of the resulting network ensemble changes continuously with data load, further improving with more training data, in analogy with the emergent behavior of large language models. Therefore, our model provides a starting point to investigate the connection among brain computation, next-token prediction, and general intelligence.

Capitalizing the future: opportunity capital as symbolic significance of an entrepreneur’s future-venture story

ABSTRACT This paper explores how a future – as something that can only be imagined but can inspire entrepreneurial action – can attract economic capital. We integrate Dor’s (2015) model of language as a communication technology for the instruction of imagination with Bourdieu’s theory of practice to account for how an entrepreneur’s words can hold sway over an investor by virtue of meaning something significant. We introduce opportunity capital as the symbolic significance of the entrepreneur’s future-venture story and the exchange counterparty to the investor’s economic capital. Our portrayal of the future as a symbolic phenomenon inducing action by others offers foundational insights to research on the early stages of venturing and the acquisition of resources by entrepreneurs.

Language in Bioethics: Beyond the Representational View

Though assumptions about language underlie all bioethical work, the field has rarely partaken of theories of language. This article encourages a more linguistically engaged bioethics. We describe the tacit conception of language that is frequently upheld in bioethics—what we call the representational view, which sees language essentially as a means of description. We examine how this view has routed the field’s theories and interventions down certain paths. We present an alternative model of language—the pragmatic view—and explore how it expands and clarifies traditional bioethical concerns. To lend concreteness, we apply the pragmatic view to a pervasive concept in bioethics and adjacent fields: decision making. We suggest that problems of the decision-making approach to bioethical issues are grounded in adherence to the representational view. Drawing on empirical work in surgery and critical care, we show how the pragmatic view productively reframes bioethical questions about how medical treatments are pursued.