Search results

The purpose of this paper is to describe a new approach to sentence representation learning leading to text classification using Bidirectional Encoder Representations from Transformers (BERT) embeddings. This work proposes a novel BERT-convolutional neural network (CNN)-based model for sentence representation learning and text classification. The proposed model can be used by industries that work in the area of classification of similarity scores between the texts and sentiments and opinion analysis.

The approach developed is based on the use of the BERT model to provide distinct features from its transformer encoder layers to the CNNs to achieve multi-layer feature fusion. To achieve multi-layer feature fusion, the distinct feature vectors of the last three layers of the BERT are passed to three separate CNN layers to generate a rich feature representation that can be used for extracting the keywords in the sentences. For sentence representation learning and text classification, the proposed model is trained and tested on the Stanford Sentiment Treebank-2 (SST-2) data set for sentiment analysis and the Quora Question Pair (QQP) data set for sentence classification. To obtain benchmark results, a selective training approach has been applied with the proposed model.

On the SST-2 data set, the proposed model achieved an accuracy of 92.90%, whereas, on the QQP data set, it achieved an accuracy of 91.51%. For other evaluation metrics such as precision, recall and F1 Score, the results obtained are overwhelming. The results with the proposed model are 1.17%–1.2% better as compared to the original BERT model on the SST-2 and QQP data sets.

The novelty of the proposed model lies in the multi-layer feature fusion between the last three layers of the BERT model with CNN layers and the selective training approach based on gated pruning to achieve benchmark results.

The purpose of this study is to examine public perceptions and attitudes toward using artificial intelligence (AI) in the US criminal justice system.

The authors took a quantitative approach and administered an online survey using the Amazon Mechanical Turk platform. The instrument was developed by integrating prior literature to create multiple scales for measuring public perceptions and attitudes.

The findings suggest that despite the various attempts, there are still significant perceptions of sociodemographic bias in the criminal justice system and technology alone cannot alleviate them. However, AI can assist judges in making fairer and more objective decisions by using triangulation – offering additional data points to offset individual biases.

Other scholars can build upon the findings and extend the work to shed more light on some problems of growing concern for society – bias and inequality in criminal sentencing. AI can be a valuable tool to assist judges in the decision-making process by offering diverse viewpoints. Furthermore, the authors bridge the gap between the fields of technology and criminal justice and demonstrate how the two can be successfully integrated for the benefit of society.

To the best of the authors’ knowledge, this is among the first studies to examine a complex societal problem like the introduction of technology in a high-stakes environment – the US criminal justice system. Understanding how AI is perceived by society is necessary to develop more transparent and unbiased algorithms for assisting judges in making fair and equitable sentencing decisions. In addition, the authors developed and validated a new scale that can be used to further examine this novel approach to criminal sentencing in the future.

The chapter is based on studies that explored the consideration of motherhood as a mitigating factor when sentencing women to imprisonment in criminal courts in England and Wales. Drawing upon two studies I conducted, it examines the way in which motherhood has been constructed by the courts and the difficulties in centring women's experiences within sentencing. It demonstrates how only by moving to a focus on children's rights has it been possible to ensure sentencers consider the implications of a defendant's motherhood.

The chapter outlines the parameters within which a sentencer can consider motherhood and explores some of the findings from the qualitative research I undertook with Crown Court judges, as part of a larger study in which I also interviewed children whose mother was at the time of interview in prison and the adults providing care to the children in her absence. The studies found that there was inconsistent understanding of a sentencer's duty to consider the impact of any sentence upon dependent children, coupled with a lack of understanding of the consequences of maternal imprisonment for a child. The ways in which this contravenes the Human Rights Act, 1998 and the United Nations Convention on the Rights of the Child 1989 are identified.

The resources which have been developed for sentencing professionals as a consequence of that work are discussed before considering the lack of concern for women's motherhood when imprisoning women on remand or recall. The chapter concludes with a call for motherhood, and not just children's rights, to be centred in sentencing decisions.

The very contextual nature of most mitigating evidence runs counter to America’s individualistic culture. Prior research has found that capital jurors are unreceptive to most mitigating circumstances, but no research has examined the capital sentencing decisions of trial judges. This study fills that gap through a content analysis of eight judicial sentencing opinions from Delaware. The findings indicate that judges typically dismiss contextualizing evidence in their sentencing opinions and instead focus predominately on the defendant’s culpability. This finding calls into question the ability of guided discretion statutes to ensure the consideration of mitigation and limit arbitrariness in the death penalty.

This chapter presents four theories that hypothesize race/ethnicity disparities in sentence outcomes. Empirical studies assessing the relationship between defendant’s race/ethnicity and sentence severity are discussed.

I focus on federal sentencing in terms of support or non-support of the theoretical perspectives.

Sentence disparity linked to defendant’s race/ethnicity are observed as net main effects, as a component in joint-conditioning effects with other extralegal defendant characteristics, and as a variable that conditions the effect of process-related mechanism of discretion, and legally relevant case characteristics, and as indirect effects.

Theories share substantial conceptual overlap in specifying the relationship between defendant’s race/ethnicity and predictions of the effect of defendant’s race/ethnicity on sentence severity.

The ways in which battered women respond to domestic violence, and the ways the legal system constructs those responses, constitute the framework of this chapter. The analysis focuses on mitigation in sentences of battered women who killed their abusers and examines the manifestation of agency and victimization in the mitigation structure. My thesis is that these women are perceived by courts solely as victims who lack agency and autonomy. Three main themes emerge from the analysis: first, the courts focus on the mental state of the defendants, stressing their psychological deficiencies as the primary mitigating factors. Secondly, many cases are categorized by courts as unique cases. Thirdly, in several cases the courts portray the women as “victims of circumstances”. An alternative analysis to that offered by the courts, one that seeks to reframe the mitigation process, is introduced in this chapter. According to this analysis, the narrative used in cases of battered women who kill should be changed to reflect dimensions of agency and resistance. In the suggested discourse, the abuse these women suffer is acknowledged, but is used to explain the women's urge to self-preservation and thus, the rationality and reasonableness of their acts.

This chapter aims to demonstrate that the fundamental human rights principle that no one should be subjected to (grossly) disproportionate punishment should be interpreted to take into account terminal illness of the offender. It should be applied both during imposition of the sentences and also during execution of already imposed sentences.

In order to reveal whether this principle takes into account serious medical conditions, including terminal illness of the offender in the calculus of the proportionality of punishment and whether it is applicable at the execution stage of sentences, this chapter examined the roots of the fundamental human rights principle of proportionality of punishment by briefly surveying the penal theory, jurisprudence, court cases, laws, and legislative history from the U.S. federal and state jurisdictions and from Europe.

There is a consensus among surveyed theories that terminal illness of the offender is an element of the principle of proportionality of punishment. Thus the fundamental human rights principle must be interpreted to take it into account. The principle should be observed not only at the imposition stage, but also at the execution stage of already imposed sentences.

This chapter re-examines the roots of the fundamental human right to not being subjected to (grossly) disproportionate punishment. It does so in order to demonstrate that the right should be interpreted to take into account terminal illness of the offender and that it should be observed not only at the imposition stage, but also at the execution stage of already imposed sentences.

This chapter explores some of the risks and constraints associated with defendants’ apologies during allocution at sentencing. It argues that defendants’ stigmatized institutional role identities in conjunction with the constraints imposed by the discursive context of allocution function to limit both the effectiveness with which defendants can speak on their own behalf and the kinds of things that they can say. Allocution has long been understood as a protection for defendants. This chapter proposes that the ideologies associated with this turn at talk have functioned to obscure the ways in which allocution preserves existing power configurations instead of challenging them.