Search results

In the COVID-19 era, sign language (SL) translation has gained attention in online learning, which evaluates the physical gestures of each student and bridges the communication gap between dysphonia and hearing people. The purpose of this paper is to devote the alignment between SL sequence and nature language sequence with high translation performance.

SL can be characterized as joint/bone location information in two-dimensional space over time, forming skeleton sequences. To encode joint, bone and their motion information, we propose a multistream hierarchy network (MHN) along with a vocab prediction network (VPN) and a joint network (JN) with the recurrent neural network transducer. The JN is used to concatenate the sequences encoded by the MHN and VPN and learn their sequence alignments.

We verify the effectiveness of the proposed approach and provide experimental results on three large-scale datasets, which show that translation accuracy is 94.96, 54.52, and 92.88 per cent, and the inference time is 18 and 1.7 times faster than listen-attend-spell network (LAS) and visual hierarchy to lexical sequence network (H2SNet) , respectively.

In this paper, we propose a novel framework that can fuse multimodal input (i.e. joint, bone and their motion stream) and align input streams with nature language. Moreover, the provided framework is improved by the different properties of MHN, VPN and JN. Experimental results on the three datasets demonstrate that our approaches outperform the state-of-the-art methods in terms of translation accuracy and speed.

Drug sales facilitated through digital communication on the surface web and on darknet cryptomarkets have increased during the past two decades. This has resulted in an increase in drug law enforcement efforts to combat these markets and a subsequent increase in judicial sentencing of people selling drugs online. The aim of this study was to analyze how Swedish courts describe sentenced sellers and how the courts apply case law.

The empirical material consists of 71 sentencing documents produced by Swedish courts in cases of online drug selling between January 1, 2010 and January 1, 2020. In total, 99 sentenced persons occur in the documents. Using a qualitative research design, the authors analyzed the material through thematic text analysis.

Overall, in their descriptions of online drug sale operations, the courts’ characterizations of the concepts of street capital and digital capital show a dichotomy. These forms of capital are situationally described as both aggravating and mitigating aspects in the application of case law, indicating that it may be fruitful to view both street and digital capital as resources used on contemporary drug markets in general.

Very little research exists into how judicial systems describe and perceive the developing phenomenon of online drug sales. Using a relatively large sample from a decade of sentencing, the authors provide an analysis of how Swedish courts view and valuate capital forms in the online drugs trade.

This paper aims to demonstrate how skills taxonomies can be used in combination with machine learning to integrate diverse online datasets and reveal skills gaps. The purpose of this study is then to show how the skills gaps revealed by the integrated datasets can be used to achieve better labour market alignment, keep educational offerings up to date and assist graduates to communicate the value of their qualifications.

Using the ESCO taxonomy and natural language processing, this study captures skills data from three types of online data (job ads, course descriptions and resumes), allowing us to compare demand for skills and supply of skills for three different occupations.

This study illustrates three practical applications for the integrated data, showing how they can be used to help workers who are disrupted by technology to identify alternative career pathways, assist educators to identify gaps in their course offerings and support students to communicate the value of their training to employers.

This study builds upon existing applications of machine learning (detecting skills from a single dataset) by using the skills taxonomy to integrate three datasets. This study shows how these complementary, big datasets can be integrated to support greater alignment between the needs and offerings of educators, employers and job seekers.

Intent detection (ID) and slot filling (SF) are two important tasks in natural language understanding. ID is to identify the main intent of a paragraph of text. The goal of SF is to extract the information that is important to the intent from the input sentence. However, most of the existing methods use sentence-level intention recognition, which has the risk of error propagation, and the relationship between intention recognition and SF is not explicitly modeled. Aiming at this problem, this paper proposes a collaborative model of ID and SF for intelligent spoken language understanding called ID-SF-Fusion.

ID-SF-Fusion uses Bidirectional Encoder Representation from Transformers (BERT) and Bidirectional Long Short-Term Memory (BiLSTM) to extract effective word embedding and context vectors containing the whole sentence information respectively. Fusion layer is used to provide intent–slot fusion information for SF task. In this way, the relationship between ID and SF task is fully explicitly modeled. This layer takes the result of ID and slot context vectors as input to obtain the fusion information which contains both ID result and slot information. Meanwhile, to further reduce error propagation, we use word-level ID for the ID-SF-Fusion model. Finally, two tasks of ID and SF are realized by joint optimization training.

We conducted experiments on two public datasets, Airline Travel Information Systems (ATIS) and Snips. The results show that the Intent ACC score and Slot F1 score of ID-SF-Fusion on ATIS and Snips are 98.0 per cent and 95.8 per cent, respectively, and the two indicators on Snips dataset are 98.6 per cent and 96.7 per cent, respectively. These models are superior to slot-gated, SF-ID NetWork, stack-Prop and other models. In addition, ablation experiments were performed to further analyze and discuss the proposed model.

This paper uses word-level intent recognition and introduces intent information into the SF process, which is a significant improvement on both data sets.

According to the Indian Sign Language Research and Training Centre (ISLRTC), India has approximately 300 certified human interpreters to help people with hearing loss. This paper aims to address the issue of Indian Sign Language (ISL) sentence recognition and translation into semantically equivalent English text in a signer-independent mode.

This study presents an approach that translates ISL sentences into English text using the MobileNetV2 model and Neural Machine Translation (NMT). The authors have created an ISL corpus from the Brown corpus using ISL grammar rules to perform machine translation. The authors’ approach converts ISL videos of the newly created dataset into ISL gloss sequences using the MobileNetV2 model and the recognized ISL gloss sequence is then fed to a machine translation module that generates an English sentence for each ISL sentence.

As per the experimental results, pretrained MobileNetV2 model was proven the best-suited model for the recognition of ISL sentences and NMT provided better results than Statistical Machine Translation (SMT) to convert ISL text into English text. The automatic and human evaluation of the proposed approach yielded accuracies of 83.3 and 86.1%, respectively.

It can be seen that the neural machine translation systems produced translations with repetitions of other translated words, strange translations when the total number of words per sentence is increased and one or more unexpected terms that had no relation to the source text on occasion. The most common type of error is the mistranslation of places, numbers and dates. Although this has little effect on the overall structure of the translated sentence, it indicates that the embedding learned for these few words could be improved.

Sign language recognition and translation is a crucial step toward improving communication between the deaf and the rest of society. Because of the shortage of human interpreters, an alternative approach is desired to help people achieve smooth communication with the Deaf. To motivate research in this field, the authors generated an ISL corpus of 13,720 sentences and a video dataset of 47,880 ISL videos. As there is no public dataset available for ISl videos incorporating signs released by ISLRTC, the authors created a new video dataset and ISL corpus.

Background: Commodity-driven deforestation is a major driver of forest loss worldwide, and globalisation has increased the disconnect between producer and consumer countries. Recent due-diligence legislation aiming to improve supply chain sustainability covers major forest-risk commodities. However, the evidence base for specific commodities included within policy needs assessing to ensure effective reduction of embedded deforestation.

Methods: We conducted a rapid evidence synthesis in October 2020 using three databases; Google Scholar, Web of Science, and Scopus, to assess the literature and identify commodities with the highest deforestation risk linked to UK imports. Inclusion criteria include publication in the past 10 years and studies that didn't link commodity consumption to impacts or to the UK were excluded. The development of a review protocol was used to minimise bias and critical appraisal of underlying data and methods in studies was conducted in order to assess the uncertainties around results.

Results: From a total of 318 results, 17 studies were included in the final synthesis. These studies used various methodologies and input data, yet there is broad alignment on commodities, confirming that those included in due diligence legislation have a high deforestation risk. Soy, palm oil, and beef were identified as critical, with their production being concentrated in just a few global locations. However, there are also emerging commodities that have a high deforestation risk but are not included in legislation, such as sugar and coffee. These commodities are much less extensively studied in the literature and may warrant further research and consideration.

Conclusion: Policy recommendations in the selected studies suggests further strengthening of the UK due diligence legislation is needed. In particular, the provision of incentives for uptake of policies and wider stakeholder engagement, as well as continual review of commodities included to ensure a reduction in the UK's overseas deforestation footprint.

Following the ‘Sarkozy’ era (2007–2012), France has engaged in ‘zero-tolerance’ policies, which have brought an increasing number of people into the criminal justice system (CJS). In an already extremely impoverished CJS, these policies have led to serious financial problems and have made an already existing prison overcrowding problem worse. Consequently, the CJS has gradually opted for a McDonald (Ritzer, 2019; Robinson, 2019) type of offender processing, whether in prosecutor-led procedures (representing roughly half of all penal procedures: Ministry of Justice, 2019) or in the sentencing phase (Danet, 2013). A similar trend has been found in probation and in prisoner release (in French: ‘sentences’ management).

The prison and probation services, which merged in 1999, have since then been in a position to benefit from the 1958 French Republic Constitution, which places the executive in a dominant position and notably allows it to draft the bills presented to a rather passive legislative power (Rousseau, 2007) and even to enjoy its own set of normative powers (‘autonomous decrees’ – Hamon & Troper, 2019). By way of law reforming (2009, 2014, and 2019 laws), the prison and probation services have thus embraced the McDonaldisation ethos. Their main obsession has been to early release as many prisoners as possible in order to free space and to accommodate more sentenced people. To do so, the prison services have created a series of so-called ‘simplified’ early release procedures, where prisoners are neither prepared for nor supported through release, where they are deprived of agency and where due process and attorney advice are removed. Behind a pretend rehabilitative discourse, the executive is only interested in efficiently flushing people out of prison; not about re-entry efficacy. As Ritzer (2019) points out, McDonaldisation often leads to counter-productive or absurd consequences. In the case of early release, the stubborn reality is that one cannot bypass actually doing the rehabilitative and re-entry work. I shall additionally argue that not everything truly qualifies as an early release measure (Ostermann, 2013). Only measures which respect prisoners’ agency prepare them for their release, which support them once they are in the community, which address their socio-psychological and criminogenic needs, and which are pronounced in the context of due process and defence rights truly qualify as such. As it is, French ‘simplified’ release procedures amount to McRe-entry and mass nothingness.

Needs for information literacy, disparities in society, bridging digital divides, richness of information sources in electronic (e-)environments and the value of dictionaries have often been propagated. To improve information sources and information literacy training, information behaviour must be understood (i.e. all information activities). This paper conceptualises new opportunities for information sources (e.g. electronic dictionaries) to all society sectors, dictionary literacy and research lenses such as lexicography to supplement information literacy and behaviour research.

A scoping review of information literacy and behaviour, lexicography and dictionary literature grounds the conceptualisation of dictionary literacy, its alignment with information literacy, information activities and information behaviour and lexicography as additional research lens.

Research lenses must acknowledge dictionary use in e-environments, information activities and skills, meanings of information and dictionary literacy, the value of e-dictionaries, alignment with information behaviour research that guides the development of information sources and interdisciplinary research from, e.g. lexicography – thus contextualisation.

Research implications – information behaviour and information literacy research can be enriched by lexicography as research lens. Further conceptualisation could align information behaviour, information literacy and dictionary literacy.

Dictionary training, aligned with information literacy training, can be informed by this paper.

The value of dictionary literacy for all sectors of societies can be improved.

Large bodies of literature on information behaviour and lexicography individually do not cover combined insights from both.

Whether automatically generated summaries of health social media can assist users in appropriately managing their diseases and ensuring better communication with health professionals becomes an important issue. This paper aims to develop a novel deep learning-based summarization approach for obtaining the most informative summaries from online patient reviews accurately and effectively.

This paper proposes a framework to generate summaries that integrates a domain-specific pre-trained embedding model and a deep neural extractive summary approach by considering content features, text sentiment, review influence and readability features. Representative health-related summaries were identified, and user judgements were analysed.

Experimental results on the three real-world health forum data sets indicate that awarding sentences without incorporating all the adopted features leads to declining summarization performance. The proposed summarizer significantly outperformed the comparison baseline. User judgement through the questionnaire provides realistic and concrete evidence of crucial features that remarkably influence patient forum review summaries.

This study contributes to health analytics and management literature by exploring users’ expressions and opinions through the health deep learning summarization model. The research also developed an innovative mindset to design summarization weighting methods from user-created content on health topics.