Monolingual Corpora Research Articles

Sarcasm Detection Based on Sentiment Analysis of Audio Corpus Using Deep Learning

With the rapid progress in digitalization, hundreds of services and applications have begun to depend on human-machine communication, primarily voice-based. As technology advances, it has become increasingly imperative to develop systems that can discern sophisticated emotions from humans, like sarcasm. Despite research that has been conducted on the detection and analysis of speech signals to try and understand emotion, the detection of sarcasm, which is often not overtly expressed, is quite far from something that has received the attention it deserves in recent years. It plays a very important role in any utterance, searching to identify the psychology, mood, or even health behind it. Hence, it does give interactions with humans a more human-like flavor while raising the feeling of understanding human emotions. This article tries to identify sarcasm in speech using audio data derived from real-world, spontaneous, monolingual corpora. In this classification process, a deep learning classifier known as CNN-LSTM model is used. Results of the study showed that a robust combination between feature extraction through convolutional layers and sequential learning through LSTM layers can identify intricate speech patterns in sarcasm. Experimented and validated in terms of performance, this model shows the model to be very good at discriminating between sarcastic and no sarcastic speech, thus promoting it for more extensive use in various applications related to sentiment analysis and in human-computer interaction applications. The results show that this system provides far much closer approximations to human interaction styles, but its impact would extend into a wide area from customer services, monitoring mental health, and even AI-driven communication systems. It is a monolingual and spontaneous corpus that will open the gates for further work and development down the line and possibly extend to multilingual and more diversified speech contexts, thus fine-tuning accuracy in practical scenarios of sarcasm detection.

Bilingual phrase induction with local hard negative sampling

AbstractBilingual lexicon induction focuses on learning word translation pairs, also known as bitexts, from monolingual corpora by establishing a mapping between the source and target embedding spaces. Despite recent advancements, bilingual lexicon induction is limited to inducing bitexts consisting of individual words, lacking the ability to handle semantics‐rich phrases. To bridge this gap and support downstream cross‐lingual tasks, it is practical to develop a method for bilingual phrase induction that extracts bilingual phrase pairs from monolingual corpora without relying on cross‐lingual knowledge. In this paper, the authors propose a novel phrase embedding training method based on the skip‐gram structure. Specifically, a local hard negative sampling strategy that utilises negative samples of central tokens in sliding windows to enhance phrase embedding learning is introduced. The proposed method achieves competitive or superior performance compared to baseline approaches, with exceptional results recorded for distant languages. Additionally, we develop a phrase representation learning method that leverages multilingual pre‐trained language models. These mPLMs‐based representations can be combined with the above‐mentioned static phrase embeddings to further improve the accuracy of the bilingual phrase induction task. We manually construct a dataset of bilingual phrase pairs and integrate it with MUSE to facilitate the bilingual phrase induction task.

Efficient incremental training using a novel NMT-SMT hybrid framework for translation of low-resource languages.

The data-hungry statistical machine translation (SMT) and neural machine translation (NMT) models offer state-of-the-art results for languages with abundant data resources. However, extensive research is imperative to make these models perform equally well for low-resource languages. This paper proposes a novel approach to integrate the best features of the NMT and SMT systems for improved translation performance of low-resource English-Tamil language pair. The suboptimal NMT model trained with the small parallel corpus translates the monolingual corpus and selects only the best translations, to retrain itself in the next iteration. The proposed method employs the SMT phrase-pair table to determine the best translations, based on the maximum match between the words of the phrase-pair dictionary and each of the individual translations. This repeating cycle of translation and retraining generates a large quasi-parallel corpus, thus making the NMT model more powerful. SMT-integrated incremental training demonstrates a substantial difference in translation performance as compared to the existing approaches for incremental training. The model is strengthened further by adopting a beam search decoding strategy to produce k best possible translations for each input sentence. Empirical findings prove that the proposed model with BLEU scores of 19.56 and 23.49 outperforms the baseline NMT with scores 11.06 and 17.06 for Eng-to-Tam and Tam-to-Eng translations, respectively. METEOR score evaluation further corroborates these results, proving the supremacy of the proposed model.

Exploring the interface between corpus and statistical modelling: a multi-methodological analysis of lexical variation in Islamic legal discourse

This research sets out to investigate how ‘culture-specific’ or ‘signature’ concepts are rendered in English-language discourse on Islamic, or ‘shariʿa law. Islamic jurisprudence is born from the fusion of Islamic scriptures and Arabic culture; therefore, it provides ample room for accommodating a distinctive legal diction that is both culture-bound and system-bound. Accordingly, the expression of shariʿa law in English would fall under the paradigm of intersystemic translation, which occurs upon the mediation between two legal systems and poses the challenge of incongruity of legal terminology. This study is based on the self-built Islamic Law Corpus (ILC), a 9-million-word monolingual English corpus, covering diverse genres on Islamic finance and family law. Using the power of corpus-processing and statistical regression software (Sketch Engine & MATLAB), this research presents a multi-methodological, multifactorial, and interdisciplinary approach to investigate lexical variation in expressing cultural phenomena within Islamic legal discourse. The analysis confirms that the contextual factors (‘genre’, ‘function’, and ‘subject field’) do influence the choice of an Arabic loanword versus English alternatives.

Telugu language hate speech detection using deep learning transformer models: Corpus generation and evaluation

In today's digital era, social media has become a new tool for communication and sharing information, with the availability of high-speed internet it tends to reach the masses much faster. Lack of regulations and ethics have made advancement in the proliferation of abusive language and hate speech has become a growing concern on social media platforms in the form of posts, replies, and comments towards individuals, groups, religions, and communities. However, the process of classification of hate speech manually on online platforms is cumbersome and impractical due to the excessive amount of data being generated. Therefore, it is crucial to automatically filter online content to identify and eliminate hate speech from social media. Widely spoken resource-rich languages like English have driven the research and achieved the desired result due to the accessibility of large corpora, annotated datasets, and tools. Resource-constrained languages are not able to achieve the benefits of advancement due to a lack of data corpus and annotated datasets. India has diverse languages that change with demographics and languages that have limited data availability and semantic differences. Telugu is one of the low-resource Dravidian languages spoken in the southern part of India.In this paper, we present a monolingual Telugu corpus consisting of tweets posted on Twitter annotated with hate and non-hate labels and experiments to provide a comparison of state-of-the-art fine-tuned deep learning models (mBERT, DistilBERT, IndicBERT, NLLB, Muril, RNN+LSTM, XLM-RoBERTa, and Indic-Bart). Through transfer learning and hyperparameter tuning, the models are compared for their effectiveness in classifying hate speech in Telugu text. The fine-tuned mBERT model outperformed all other fine-tuned models achieving an accuracy of 98.2. The authors also propose a deployment model for social media accounts.

Creation of a diverse mixed-lingual emotional speech corpus with a framework for enhanced emotion detection

Speech Emotion Recognition (SER) has advanced considerably during the past 20 years. Till date, various SER systems have been developed for monolingual, multilingual and cross corpus contexts. However, in a country like India where numerous languages are spoken and often humans converse in more than one language, a dedicated SER system for mixed-lingual scenario is more crucial to be established which is the focus of this work. A self-recorded database that includes speech emotion samples with 11 diverse Indian languages has been developed. In parallel, a mixed-lingual database is formed with three popular standard databases of Berlin, Baum and SAVEE to represent mixed-lingual environment for western background. A detailed investigation of GeMAPS (Geneva Minimalistic Acoustic Parameter Set) feature set for mixed-lingual SER is performed. A distinct set of MFCC (Mel Frequency Cepstral Coefficients) coefficients derived from sine and cosine-based filter banks enriches the GeMAPS feature set and are proven to be robust for mixed-lingual emotion recognition. Various Machine Learning (ML) and Deep Learning (DL) algorithms have been applied for emotion recognition. The experimental results demonstrate GeMAPS features classified from ML has been quite robust for recognizing all the emotions across the mixed-lingual database of the western languages. However, with diverse recording conditions and languages of the Indian self-recorded database the GeMAPS with enriched features and classified using DL are proven to be significant for mixed-lingual emotion recognition.

Improving linear orthogonal mapping based cross-lingual representation using ridge regression and graph centrality

Orthogonal linear mapping is a commonly used approach for generating cross-lingual embedding between two monolingual corpora that uses a word frequency-based seed dictionary alignment approach. While this approach is found to be effective for isomorphic language pairs, they do not perform well for distant language pairs with different sentence structures and morphological properties. For a distance language pair, the existing frequency-aligned orthogonal mapping methods suffer from two problems - (i)the frequency of source and target word are not comparable, and (ii)different word pairs in the seed dictionary may have different contribution. Motivated by the above two concerns, this paper proposes a novel centrality-aligned ridge regression-based orthogonal mapping. The proposed method uses centrality-based alignment for seed dictionary selection and ridge regression framework for incorporating influential weights of different word pairs in the seed dictionary. From various experimental observations over five language pairs (both isomorphic and distant languages), it is evident that the proposed method outperforms baseline methods in the Bilingual Dictionary Induction(BDI) task, Sentence Retrieval Task(SRT), and Machine Translation. Further, several analyses are also included to support the proposed method.

A study on the innovative model of foreign language teaching in universities using big data corpus

This paper aims to explore unsupervised cross-lingual word representation learning methods with the specific task of acquiring a bilingual translation lexicon on a monolingual corpus. Specifically, an unsupervised cross-lingual word representation co-training scheme based on different word embedding models is first designed and outperforms the baseline model. In this paper, we adeptly tackles the obstacles encountered in higher education foreign language teaching and underscores the necessity for inventive teaching methods, and design and implement a linear self-encoder-based principal component acquisition scheme for the interpoint mutual information matrix obtained from a monolingual corpus. And on top of this, a collaborative training scheme based on linear self-encoder for cross-language word representation is designed to improve the learning effect of cross-language word embedding. The results of the study show that the most obvious rise in the pre and post tests of the experimental class in the practical application of the foreign language teaching model based on the method of this paper is the word sense guessing, which rose by 23.12%. Sentence meaning comprehension increased by 23.39%, main idea by 16.61%, factual details by 15.47%, and inferential judgment by 10.28%. Thus, the feasibility of the unsupervised cross-linguistic word representation learning collaborative training method is further verified.

Consensus-Based Machine Translation for Code-Mixed Texts

Multilingualism in India is widespread due to its long history of foreign acquaintances. This leads to the presence of an audience familiar with conversing using more than one language. Additionally, due to the social media boom, the usage of multiple languages to communicate has become extensive. Hence, the need for a translation system that can serve the novice and monolingual user is the need of the hour. Such translation systems can be developed by methods such as statistical machine translation and neural machine translation, where each approach has its advantages as well as disadvantages. In addition, the parallel corpus needed to build a translation system, with code-mixed data, is not readily available. In the present work, we present two translation frameworks that can leverage the individual advantages of these pre-existing approaches by building an ensemble model that takes a consensus of the final outputs of the preceding approaches and generates the target output. The developed models were used for translating English-Bengali code-mixed data (written in Roman script) into their equivalent monolingual Bengali instances. A code-mixed to monolingual parallel corpus was also developed to train the preceding systems. Empirical results show improved BLEU and TER scores of 17.23 and 53.18 and 19.12 and 51.29, respectively, for the developed frameworks.

Epistemic modality in English and Thai medical abstracts: a corpus-based study

ABSTRACT Epistemicity is roughly understood as the linguistic expression of a speaker’s commitment to what they say or write. Against a backdrop of scant knowledge about how the phenomenon works in many non-European languages, this study seeks to explore epistemicity in translated and original Thai. Two specialized corpora were built. The first, a parallel corpus of medical abstracts translated from English into Thai (about 100,000 English and 150,000 Thai words), was used to investigate translated Thai and its relationship with the English source texts. The second, a 250,000-word monolingual corpus of medical abstracts originally written in Thai, was used to compare translated and non-translated texts. The distribution and strength of epistemic modal expressions in translated Thai abstracts were found to follow closely patterns established in the English source texts. They differed significantly from patterns observed in original Thai, however, leading to the tentative conclusion that translated Thai could impart less certainty on the writers’ side than would normally be perceived by readers of medical abstracts originally written in Thai.

A cross-guidance cross-lingual model on generated parallel corpus for classical Chinese machine reading comprehension

Chinese diachronic gap is a key issue in classical Chinese machine reading comprehension (CCMRC). Preceding work on bridging this gap has been mostly restricted to limited monolingual classical Chinese corpora pre-training and lexical knowledge integration, which require a great deal of human resources. In this paper, we propose a cross-guidance cross-lingual model (CGCLM), pre-trained on a classical and modern Chinese parallel corpus generated from a large language model, to bridge the Chinese diachronic gap and reduce the manual effort. The CGCLM facilitates accurate translation by providing in-context examples and feedback based on the longest common substring between source and target sentences, thereby avoiding untranslated Chinese words. Specifically, we consider three pre-training tasks, i.e., cross-masked language modeling, linguistic label cross-prediction, and semantic cross-aware translation language modeling. The knowledge acquired from masked tokens uncovering and linguistic label predicting can lead to the implicit semantic alignment between two language styles. Taking advantage of the semantic similarity between the same syntactic levels of parallel pairs, cross-aware modeling integrates and transmits contextualized semantic information. We utilize an 18.6G monolingual corpus to create a 37.2G parallel corpus. Manual evaluation has resulted in only acceptable discrepancies between our generated and human-edited parallel corpora. Extensive experimental results show that our proposed model outperforms the state-of-the-art by an average accuracy of 3.13%, 2.44%, and 2.17% on CCMRC, classical Chinese language understanding evaluation (CCLUE), and modern Chinese language understanding evaluation (MCLUE) tasks.

Filtered data augmentation approach based on model competence evaluation

The scale of parallel corpus plays an important role in training high-quality neural machine translation models. In order to expand the scale of parallel corpus in low-resource scenarios, researchers have proposed a series of data augmentation approaches, in which the most representative work is the back-translation. The back-translation approach uses a basic translation model to translate the target monolingual corpus into the source language, and then combines it into a pseudo parallel corpus to expand the training data. Due to the simple and efficient data expansion strategies, the back-translation approach becomes the mainstream data augmentation approach of neural machine translation. However, in low-resource scenarios, only low-precision basic translation models can be used. The pseudo-parallel corpus translated by this model contains low-quality sentence pairs, which inevitably introduces noises and leads to negative impacts to translation procedure. To improve the performance of low-resource neural machine translation, we propose a filtered data augmentation method based on the model competence evaluation, which can effectively improve the quality of training data by using both the filtered back-translation and dynamic evaluation of translation model capabilities. Comparative experiments show that the proposed approach can improve the quality of training data and the performance of low-resource neural machine translation.

Fluidic motion patterns in English and Modern Greek

Abstract This paper investigates conceptual and phraseological patterns from a cross-linguistic perspective. The focus of attention is on the fluidic motion uses of the highly polysemous verbs run and τρέχω ‘run’ in English and Modern Greek respectively. They are manner of motion verbs denoting the typically human, fast movement on ground and they are frequently cited in the literature on motion event encoding; yet, their extended use to denote motion of liquids is mentioned only in passing. The study thus provides a comprehensive description of that part of the semantic network of the two verbs that relates to fluidic motion (literally and figuratively). The contrastive approach taken combines cognitive semantics (Frame Semantics, Conceptual Metaphor and Metonymy Theory) with a phraseological view of language. Convergences and divergences are identified at a conceptual and phraseological level through a twofold corpus-based study involving comparable monolingual analysis and parallel corpus investigation.

The atypical uses of causative make in Arab EFL learner writing: A corpus-based study

This study analyzed the atypical uses of make as a periphrastic causative in Arabic-speaking English as a Foreign Language (EFL) learner writing. Data from a learner corpus, The Arab College Corpus, were compared with data from The Louvain Corpus of Native English Essays, from which it was found that there was an overuse of the causative make and a preference for verb over adjective complements. The qualitative analysis revealed that the causative make sometimes substituted enabling causatives such as let or allow. Some usage patterns were attributed to proficiency limitations, and other usage patterns were related to the learner's spoken native dialect, namely Spoken Palestinian Arabic (SPA). The monolingual corpus of SPA Curras data confirmed that the causative xalla had both enabling and effectuating functions, with the two meanings forming a continuum. It is proposed that this continuum is being transferred to the causative make by Arab EFL learners.

The effect of standardisation on collocations in Colombian and European Spanish translations of the work "Rootless"

The empirical study of syntactic, stylistic, and lexical patterns through corpora has frequently revealed a tendency towards standardisation, as Toury (1995/2012) defines it, in translated texts. In this vein, some studies have found that diatopy tends to become blurred in translations, particularly into transnational languages. In the case of Spanish, it has been stated that the enormous geographical richness of the language is not really reflected in translated Spanish. This paper argues that the translation of the work Rootless by Chris Howard into Colombian Spanish exhibits traits more typical of the European Spanish than of the Colombian Spanish variety. For this purpose, diatopic distribution of verb + noun (object) collocations extracted from a parallel corpus and a monolingual corpus of Rootless and its translations published in Colombian and European Spanish, is examined in a reference Spanish corpus.
 Keywords: parallel corpus; translated Spanish; Colombian Spanish; standardisation in translation; diatopic variation

A Corpus-based Study on Chunk-Explicitation in Interpreting: A Case Study of Chinese Leaders’ Speeches under the COVID-19 Pandemic

This study investigates the phenomenon of chunk explicitation in Chinese-English diplomatic interpretation during the COVID-19 pandemic. It focuses on high-frequency chunks employed by prominent national leaders, including General Secretary Xi Jinping, Premier Li Keqiang, and State Councilor Wang Yi. The research utilizes monolingual corpus analysis and AntConc software to analyze six actual speech interpretations. It categorizes explicitation into four types: obligatory, optional, pragmatic, and translation-inherent explicitation. The study identifies three main motivations behind chunk explicitation in diplomatic interpretation: linguistic motivations stemming from syntactic and grammatical differences, cultural motivations arising from bridging cultural gaps, and interpreter motivations related to conveying information effectively. Additionally, the research assesses the implications of explicitation, including its impact on clarity, comprehensibility, and its role in conveying emotions and tone in diplomatic speeches. Chunk explicitation is a common phenomenon in Chinese-English diplomatic interpretation during the pandemic. It serves linguistic, cultural, and interpreter motivations, enhancing clarity and conveying emotions in diplomatic communication. The study contributes to diplomatic interpretation and provides practical guidance for interpreting learners and practitioners. Future research may expand the scope and classification of interpretation types, further advancing our understanding of this complex process. Ultimately, this research aids in promoting effective diplomatic communication and global understanding during crises.

Exploring the Pedagogical Potential of Monolingual Corpora in Inverse Translation Training

This article presents a comprehensive study that delves into the untapped potential of incorporating monolingual corpora as a strategic tool within translator training to enhance trainees’ proficiency in the challenging domain of inverse translation from Arabic to English. By shedding light on the pedagogical promise of monolingual corpora in inverse translation instruction, this research underscores the critical need to integrate this tool into translator training curricula, thereby amplifying the translation competencies of trainees. The central research inquiry is as follows: How does using a monolingual corpus enhance the performance of students engaged in inverse translation of collocations from Arabic to English within the context of political and media texts? The researchers focused on two groups of female students specializing in political and media translation at the College of Language Sciences, King Saud University. These groups, a control and an experimental one, were meticulously trained with different references—while the control group relied on conventional bilingual dictionaries, the experimental group employed monolingual corpora as their primary translation references—the assessment criterion centered on translating a newspaper text. The noteworthy outcome of this comparative analysis was that the experimental group consistently outperformed their counterparts in the control group by leveraging monolingual corpora. The findings from this research offer precise guidance to enhance the efficiency of translation training programs. Moreover, the study equips translators with essential skills and competencies and elevates the quality of translation training programs.

Investigating Unsupervised Neural Machine Translation for Low-resource Language Pair English-Mizo via Lexically Enhanced Pre-trained Language Models

The vast majority of languages in the world at present are considered to be low-resource languages. Since the availability of large parallel data is crucial for the success of most modern machine translation approaches, improving machine translation for low-resource languages is a key challenge. Most unsupervised techniques for translation benefit closely related languages with monolingual data of substantial quantity. To facilitate research in this direction for the extremely low resource language pair English ( en ) and Mizo ( lus ), we have developed a parallel and monolingual corpus for the Mizo language from various news websites. We explore Unsupervised Neural Machine Translation (UNMT) based on the developed monolingual data. We observe that cross-lingual embedding (CLWE) initializations on subword segmented data during pre-training, based on both masked language modelling and sequence-to-sequence generation tasks, improve translation performance. We experiment with cross-lingual alignment and combined alignment and joint training for learning the cross-lingual embedding representations. We also report baseline performances and the impact of CLWE initialization using semi-supervised and supervised neural machine translation. Empirical results show that both CLWE initializations work well for the distant pair English-Mizo compared to the baselines.

A content-controlled monolingual comparable corpus approach to comparing learner and proficient argumentative writing

This mixed-methods study approaches the differences between learner and proficient argumentative writing by building a content-controlled monolingual comparable corpus (CCMCC) that contains learner-teacher sample pairs of the same semantic content. Twenty-seven learner samples were collected from 27 Chinese university students who each wrote on one topic from the second writing task of IELTS Academic. To generate content-controlled teacher samples, an experienced teacher revised or rewrote each learner sample after confirming the ideas learners intended to express through individual and face-to-face communication with each learner. Then, a native speaker checked the language of the teacher samples. In data analysis, each learner-teacher sample pair was analyzed using Coh-Metrix to generate statistics in 45 indices under text length, syntax, lexicon, and cohesion, after which a shortlist of indices of both statistically and practically significant differences was identified. Qualitatively, the researchers identified the differences through side-by-side comparisons of sample pairs and coded the important patterns in the differences to explore their underlying reasons. This approach generated different quantitative results from previous corpus-based comparative writing studies, such as the ineffectiveness of cohesion indices to distinguish learner and proficient writing. Qualitative analysis further revealed noteworthy findings including the lack of concision in learner writing and learners’ unfamiliarity with using prepositional phrases to express actions. The advantages, limitations and implications of this approach are discussed.

Bilingual word embedding fusion for robust unsupervised bilingual lexicon induction

Great progress has been made in unsupervised bilingual lexicon induction (UBLI) by aligning the source and target word embeddings independently trained on monolingual corpora. The common assumption of most UBLI models is that the embedding spaces of two languages are approximately isomorphic (i.e., similar in geometric structure). Therefore, the performance is bound by the degree of isomorphism, especially on etymologically and typologically distant languages. Near-zero UBLI results have been reported for them. To address this problem, we propose a method to increase the isomorphism based on bilingual word embedding fusion. In particular, the features from the source embeddings are integrated into the target embeddings, and vice versa. Therefore, the resulting structures of source and target embeddings are similar to each other. The method does not require any form of supervision and can be applied to any language pair. On a benchmark dataset of bilingual lexicon induction, our approach can achieve competitive or superior performance compared to the state-of-the-art methods, with particularly strong results being found on distant languages.