12 research outputs found
A study of the translation of sentiment in user-generated text
No full textA thesis submitted in partial ful filment of the requirements of the University of Wolverhampton for the degree of Doctor of Philosophy.Emotions are biological states of feeling that humans may verbally express to
communicate their negative or positive mood, influence others, or even afflict
harm. Although emotions such as anger, happiness, affection, or fear are
supposedly universal experiences, the lingual realisation of the emotional
experience may vary in subtle ways across different languages. For this reason,
preserving the original sentiment of the source text has always been a challenging
task that draws in a translator's competence and fi nesse. In the professional
translation industry, an incorrect translation of the sentiment-carrying lexicon is
considered a critical error as it can be either misleading or in some cases harmful
since it misses the fundamental aspect of the source text, i.e. the author's
sentiment.
Since the advent of Neural Machine Translation (NMT), there has been a
tremendous improvement in the quality of automatic translation. This has lead to
an extensive use of NMT online tools to translate User-Generated Text (UGT)
such as reviews, tweets, and social media posts, where the main message is often
the author's positive or negative attitude towards an entity. In such scenarios, the
process of translating the user's sentiment is entirely automatic with no human
intervention, neither for post-editing nor for accuracy checking. However, NMT
output still lacks accuracy in some low-resource languages and sometimes makes
critical translation errors that may not only distort the sentiment but at times flips
the polarity of the source text to its exact opposite.
In this thesis, we tackle the translation of sentiment in UGT by NMT systems from two perspectives: analytical and experimental. First, the analytical approach
introduces a list of linguistic features that can lead to a mistranslation of
ne-grained emotions between different language pairs in the UGT domain. It also
presents an error-typology specifi c to Arabic UGT illustrating the main linguistic
phenomena that can cause mistranslation of sentiment polarity when translating
Arabic UGT into English by NMT systems. Second, the experimental approach
attempts to improve the translation of sentiment by addressing some of the
linguistic challenges identifi ed in the analysis as causing mistranslation of
sentiment both on the word-level and on the sentence-level. On the word-level, we
propose a Transformer NMT model trained on a sentiment-oriented vector space
model (VSM) of UGT data that is capable of translating the correct sentiment
polarity of challenging contronyms. On the sentence-level, we propose a
semi-supervised approach to overcome the problem of translating sentiment
expressed by dialectical language in UGT data. We take the translation of
dialectical Arabic UGT into English as a case study. Our semi-supervised AR-EN
NMT model shows improved performance over the online MT Twitter tool in
translating dialectical Arabic UGT not only in terms of translation quality but
also in the preservation of the sentiment polarity of the source text. The
experimental section also presents an empirical method to quantify the notion of
sentiment transfer by an MT system and, more concretely, to modify automatic
metrics such that its MT ranking comes closer to a human judgement of a poor or
good translation of sentiment
Sentiment-Aware Measure (SAM) for Evaluating Sentiment Transfer by Machine Translation Systems
No full textIn translating text where sentiment is the main message, human translators give particular attention to sentiment-carrying words. The reason is that an incorrect translation of such words would miss the fundamental aspect of the source text, i.e. the author’s sentiment. In the online world, MT systems are extensively used to translate User-Generated Content (UGC) such as reviews, tweets, and social media posts, where the main message is often the author’s positive or negative attitude towards the topic of the text. It is important in such scenarios to accurately measure how far an MT system can be a reliable real-life utility in transferring the correct affect message. This paper tackles an under-recognised problem in the field of machine translation evaluation which is judging to what extent automatic metrics concur with the gold standard of human evaluation for a correct translation of sentiment. We evaluate the efficacy of conventional quality metrics in spotting a mistranslation of sentiment, especially when it is the sole error in the MT output. We propose a numerical ‘sentiment-closeness’ measure appropriate for assessing the accuracy of a translated affect message in UGC text by an MT system. We will show that incorporating this sentiment aware measure can significantly enhance the correlation of some available quality metrics with the human judgement of an accurate translation of sentiment
Towards a Better Understanding of Tarajem: Creating Topological Networks for Arabic biographical Dictionaries
No full textBiographical writing is one of the earliest and most extensive forms of Arabic literature. Some scholars tend to assume that classical Arabic biographies, widely known as Tarāǧim, arose in conjunction with the study of the reliability of the Hadith transmitters (the reciters of the Prophet Mohammad's sayings) which lead to a proliferation of biographical material collected and used to assess the transmitter's trustworthiness . However, a scrutiny of the well-known classical Arabic biographical dictionaries such as Siyaru 'A`lāmi an-Nubalā' `The Lives of the Noble Figures' for Adh-Dhahabī shows that they extend their entries to other classes of persons important to the development of particular fields such as Islamic jurisprudents, rulers, poets, philosophers or physicians. The main contribution of Arabic biographical dictionaries is the cumulative value of the thousands of life histories which construct a picture of the Islamic society in different eras. An Arabic biographical dictionary, therefore, is predominantly used by scholars to look up an eminent person's achievements and historical background. In this project, however, we explore Arabic biographies as a prosopography, rather than a biography in the strict sense. We introduce a novel method for a better understanding of Arabic biographical dictionaries by creating a network of relations among different persons. We utilise Natural Language Processing (NLP) tools to create a topological network from the unstructured data of 45,500 biographical entries collected from different dictionaries. We aim to illustrate how network analysis leveraged by NLP tools can provide scholars with innovative methods for discovering complex constellation of relations between prominent and non-prominent figures spanning over several eras and from different fields of knowledge. We also use graph visualisation as a means to effectively communicate and explore such complex constellations. Each network visualisation is purposefully designed to be as simple and robust as possible to offer scholars a way to move relatively fluidly between the large scale of biographical entries and to easily interpret the minute ties between persons of different walks of life. We make both our data and code publicly available for researchers to replicate the experiment. It can be found at:https://github.com/sadanyh/Relational-Network-for-Arabic-Taraje
Cyber Risks of Machine Translation Critical Errors : Arabic Mental Health Tweets as a Case Study
No full textWith the advent of Neural Machine Translation (NMT) systems, the MT output
has reached unprecedented accuracy levels which resulted in the ubiquity of MT
tools on almost all online platforms with multilingual content. However, NMT
systems, like other state-of-the-art AI generative systems, are prone to errors
that are deemed machine hallucinations. The problem with NMT hallucinations is
that they are remarkably \textit{fluent} hallucinations. Since they are trained
to produce grammatically correct utterances, NMT systems are capable of
producing mistranslations that are too fluent to be recognised by both users of
the MT tool, as well as by automatic quality metrics that are used to gauge
their performance. In this paper, we introduce an authentic dataset of machine
translation critical errors to point to the ethical and safety issues involved
in the common use of MT. The dataset comprises mistranslations of Arabic mental
health postings manually annotated with critical error types. We also show how
the commonly used quality metrics do not penalise critical errors and highlight
this as a critical issue that merits further attention from researchers
Employing AI for Better Access to Justice: An Automatic Text-to-Video Linking Tool for UK Supreme Court Hearings
No full textThe increasing adoption of artificial intelligence across domains presents new opportunities to enhance access to justice. In this paper, we introduce a human-centric AI tool that utilises advances in Automatic Speech Recognition (ASR) and Large Language Models (LLMs) to facilitate semantic linking between written UK Supreme Court (SC) judgements and their corresponding hearing videos. The motivation stems from the critical role UK SC hearings play in shaping landmark legal decisions, which often span several hours and remain difficult to navigate manually. Our approach involves two key components: (1) a customised ASR system fine-tuned on 139 h of manually edited SC hearing transcripts and legal documents and (2) a semantic linking module powered by GPT-based text embeddings adapted to the legal domain. The ASR system addresses domain-specific transcription challenges by incorporating a custom language model and legal phrase extraction techniques. The semantic linking module uses fine-tuned embeddings to match judgement paragraphs with relevant spans in the hearing transcripts. Quantitative evaluation shows that our customised ASR system improves transcription accuracy by 9% compared to generic ASR baselines. Furthermore, our adapted GPT embeddings achieve an F1 score of 0.85 in classifying relevant links between judgement text and hearing transcript segments. These results demonstrate the effectiveness of our system in streamlining access to critical legal information and supporting legal professionals in interpreting complex judicial decisions
Effect of connecting shunt capacitor on nonlinear load terminals
No full textThis material is posted here with permission of the IEEE. Such permission of the IEEE does not in any way imply IEEE endorsement of any of Brunel University's products or services. Internal or personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution must be obtained from the IEEE by writing to [email protected]. Copyright @ 2003 IEEEThe use of terminal shunt capacitance has different effects on the displacement factor and distortion factor components of the power factor. These effects are considered for nonlinear loads with ideal supply, and also where the supply impedance exists but is small compared with the load impedance. Optimization of the displacement factor is found to result in reduction of the distortion factor to a minimum value
RGCL at IDAT: deep learning models for irony detection in Arabic language
No full textThis article describes the system submitted by the RGCL team to the IDAT 2019
Shared Task: Irony Detection in Arabic Tweets. The system detects irony in Arabic tweets using
deep learning. The paper evaluates the performance of several deep learning models, as well as
how text cleaning and text pre-processing influence the accuracy of the system. Several runs
were submitted. The highest F1 score achieved for one of the submissions was 0.818 making the
team RGCL rank 4th out of 10 teams in final results. Overall, we present a system that uses
minimal pre-processing but capable of achieving competitive results
Human-in-the-Loop Learning With LLMs for Efficient RASE Tagging in Building Compliance Regulations
No full textAutomated compliance checking (ACC) in the Architecture, Engineering, and Construction (AEC) sector represents a pivotal task which is traditionally executed manually, demanding significant time and labor. This work investigates the automation of the Requirement, Applicability, Selection, and Exception (RASE) methodology for building regulatory compliance through the utilization of Large Language Models (LLMs) and active learning techniques. Specifically, we focus on the development and assessment of a system using the OpenAI GPT-4o model to transmute building regulation texts into structured YAML formats conducive to ACC processes. The study encompasses three experimental paradigms: few-shot learning, fine-tuning learning, and progressive active learning. Initial results from the few-shot learning experiment illustrate the model’s preliminary ability to interpret and process regulatory texts with limited examples. Fine-tuning enhances model performance by training it on a specialized dataset, thereby improving structural and textual accuracy. Progressive active learning, by iteratively incorporating expert feedback, further refines the accuracy of the model. The findings demonstrate substantial enhancements in both structural and semantic accuracies of the generated YAML files, underscoring the potential of integrating LLMs with active learning to streamline regulatory compliance automation. The methodologies and results presented here offer a comprehensive framework for advancing future research and practical applications in the domain of automated regulatory compliance
Google Translate Error Analysis for Mental Healthcare Information: Evaluating Accuracy, Comprehensibility, and Implications for Multilingual Healthcare Communication
No full textThis study explores the use of Google Translate (GT) for translating mental
healthcare (MHealth) information and evaluates its accuracy, comprehensibility,
and implications for multilingual healthcare communication through analysing GT
output in the MHealth domain from English to Persian, Arabic, Turkish,
Romanian, and Spanish. Two datasets comprising MHealth information from the UK
National Health Service website and information leaflets from The Royal College
of Psychiatrists were used. Native speakers of the target languages manually
assessed the GT translations, focusing on medical terminology accuracy,
comprehensibility, and critical syntactic/semantic errors. GT output analysis
revealed challenges in accurately translating medical terminology, particularly
in Arabic, Romanian, and Persian. Fluency issues were prevalent across various
languages, affecting comprehension, mainly in Arabic and Spanish. Critical
errors arose in specific contexts, such as bullet-point formatting,
specifically in Persian, Turkish, and Romanian. Although improvements are seen
in longer-text translations, there remains a need to enhance accuracy in
medical and mental health terminology and fluency, whilst also addressing
formatting issues for a more seamless user experience. The findings highlight
the need to use customised translation engines for Mhealth translation and the
challenges when relying solely on machine-translated medical content,
emphasising the crucial role of human reviewers in multilingual healthcare
communication
Mitigating mismatch power loss of series-parallel and total-cross-tied array configurations using novel enhanced heterogeneous hunger games search optimizer
No full textData availability: Data will be made available on request.Copyright © 2022 The Author(s). The location of shaded or faulty Photovoltaic modules in the PV array has a negative impact on the harvested power from the entire array. To overcome this significant limitation, PV reconfiguration is a considerable technique developed via interchanging the PV modules’ location physically or electrically. By this inspiration, in this article, the authors propose a novel enhanced heterogeneous hunger games search optimizer (EHHGS) based PV reconfiguration. The innovated EHHGS introduces a modified variant for the basic hunger game search optimizer (HGS) to achieve a high diversity and robust exploitation of the optimal solutions. The EHHGS is applied to identify the optimal relocation for the shaded or faulty modules in two configurations of PV connected array: total-cross-tied array (TCT) and Series–parallel one (S–P). The proposed approach has applied symmetric and asymmetric connected PV arrays with dimensions of 9 × 9 and 10 × 8 throughout five different shade patterns. Moreover, for providing a flexible tool for the user/researcher to detect and observe the benefits achieved via the PV reconfiguration strategy, a simple graphical user interface (GUI) for the PV reconfiguration strategy of TCT or S–P PV connected array using meta-heuristic algorithms is designed. This implemented GUI can extend for any size of PV arrays, different optimization algorithms, and different connection schemes. The proposed EHHGS, HGS, and set of recent optimizers, including harris hawk optimizer (HHO), marine predators algorithm (MPA), and artificial ecosystem-based optimization (AEO), handle a new simplified objective function to boost the optimizer’s ability in catching the optimal modules’ location to alleviate the mismatched power in the studied arrays. Several statistical metrics are computed for providing an unbiased comparison. Through the comparisons, the proposed EHHGS exhibits superior performance. It achieves the best re-design for the considered arrays that helps in avoiding the mismatch losses in the cases of the partial shaded/faulty modules and enhances the power generated profiles. EHHGS enhances the power by percentages of 44.42%, 11.9%, 33.36%, 20. 86% and 13.17% compared to the TCT-connected system. In the case of the S–P connection, the proposed EHHGS generates 47.2% and 10.45%, 30.75%, 17.25%, and 26.27% higher power.CIRA-013-2020, Khalifa University, Abu Dhabi, United Arab Emirates
