Detection and Mapping of Violent Crime Hotspots in Southwestern Nigeria

The increasing rate of insecurity in Nigeria, especially the southwest requires a paradigm shift from popular approach to crime hotspots detection. This study employed geospatial technologies to integrate spatio-temporal crime, social media and field observation data from the communities in all the six states in the southwest to develop crime hotspots that can serve as preliminary information to assist in allocating resources for crime control and prevention. Historical crime data from January 1972 to April, 2021 were compiled and updated with rigorous field survey in September, 2021. The field data were encoded, input to the SPSS 17 and analyzed using descriptive statistics and multivariate analysis. A total 936 crime locations data were geolocated and exported to ArcGIS 10.5 for spatial mapping using point map operation and further imported to e-Spatial web-based and QGIS for the generation of hotspot map using heatmap tool. The results revealed that armed robbery, assassination and cultism were more pronounced in Lagos and Ogun States. Similarly, high incidences of farmers/herdsmen conflicts are observed in Oyo and Osun States. Increasing incidences of kidnapping are common in all the south-western states but very prominent in Ondo, Lagos and Oyo States. Most of the violent crime incidents took place along the highways, with forests being their hideouts. Violent crimes are dominantly caused by high rate of unemployment while farmer/herdsmen conflicts were majorly triggered by the scarcity of grazing fields and destruction of arable crops. The conflicts have resulted in the increasing cases of rape and disruption of social group, intake of hard drugs, cult-related activities, low income and revenue generation, and displacement of farmers and infrastructural damages. The study advocates regular retraining and equipping of security agents, establishment of cattle ranch, and installation of sophisticated IP Camera at the crime hotspots to assist in real-time crime monitoring and management.

Prevalence, Pattern and Factors Associated with Psychiatric Disorders among Persons Arrested for Violent Crimes in Ife/Modakeke Area, Southwestern Nigeria

Background: Association between violence and mental disorders has contributed immensely to the stigma associated with mental illness in the society;because people erroneously see mentally ill individuals as dangerous, they will not want to associate with them. Aims: To assess the prevalence and pattern of psychiatric disorders among a sample of the violent offenders and to examine any relationship between psychiatric disorders and crimes. Method: This was a two-phase cross-sectional study in three police stations in Ile-Ife/Modakeke area of Nigeria. In the first phase, we screened 400 consecutive adults arrested for violent crimes using the General Health Questionnaire—30. In the second phase, all 36 persons with probable psychopathology were then interviewed with the Present State Examination to make a definitive diagnosis. Results: The mean age of all the subjects was 29.9 years (SD ± 7.3). The male to female ratio was 11:1. Respondents were mostly single (54%);most had secondary education or less (82%) and about 60% were currently using psychoactive substances (drugs). About 8.5% of all the subjects had a diagnosable psychiatric disorder;paranoid schizophrenia was the commonest psychiatric disorder (41.2%). Mentally ill subjects were three times more likely to commit homicidal offence than non-mentally ill subjects. Conclusion: There exists a significant but weak relationship between mental illness and violent crimes.

A New Interface Identification Technique Based on Absolute Density Gradient for Violent Flows

An identification technique for sharp interface and penetrated isolated particles is developed for simulating two-dimensional,incompressible and immiscible two-phase flows using meshless particle methods in this paper.This technique is based on the numerically computed density gradient of fluid particles and is suitable for capturing large interface deformation and even topological changes such as merging and breaking up of phases.A number of assumed particle configurations will be examined using the technique,including these with different level of randomness of particle distribution.The tests will show that the new technique can correctly identify almost all the interface and isolated particles,and also show that it is better than other existing popular methods tested.

Non-Violent Communication and Its Impact on Suicidal Ideations and Coping with Stress Styles

One of the important dimensions of violence is suicide, or violence directed toward self that many people suffer from nowadays. In the meantime, adolescents and especially girls are more likely to be harmed. The purpose of this study was to evaluate the effect of group training with non-violent communication (NVC) method on suicidal ideations and coping with stress styles of adolescent girls. The target community in this research was availably selected from the charity centers of Tehran suburbs. In this study, 15 adolescent girls were in the experimental group and 15 in the control group. The criteria for entering the research were students aged 11 to 19 years and the vulnerable socio-economic income. Eight training sessions of NVC were presented for the experimental group in the pre-test and post-test intervals, while no intervention was performed for the control group. Data collection instruments were Beck’s scale for suicidal ideation (BDI) and Endler & Parker’s coping inventory for stressful situations (CISS). After statistical analysis, the effectiveness of this training was investigated on dependent variables by multivariate analysis of covariance. Eventually, the results showed a significant difference between the experimental and control groups (p < 0.05). The finding of the study showed that NVC training can reduce suicidal ideations in adolescent girls who are socially and economically vulnerable. According to the findings, non-violent communication training has a higher effect on emotion-focused coping style compared to other stress coping styles.

The Short-Term Effect of Online Violent Stimuli on Aggression

The significance of this study was to find whether violent stimuli exposure could escalate the following levels of aggression in order to better cultivate aggression education among contemporary undergraduates in China. This study mainly tested the effects of violent stimuli on aggression by employing modified Stroop task. A total of 188 undergraduates participated in this study. Results showed that undergraduates exposing to violent stimuli exhibited high levels of aggression, whereas undergraduates who exposed to non-violent stimuli displayed low levels of aggression. Specifically, males, but not females, manifested high levels of aggression after exposure to violent stimuli. Moreover, viewers with high-aggressiveness (HA), but not with moderate-aggressiveness (MA) and low-aggressiveness (LA), showed high levels of aggression after exposure to violent stimuli via internet.

Prevention of Violent Extremism and Human Rights Protection

In the context of globalization, violent extremist forces pose a serious threat to the fundamental human rights of the people all over the world. Strengthening human rights protection requires reinforcing the momentum to crack down and prevent the violent acts of extremist forces. As one of the countries that is the victim of violence by extremists, China has taken a series of measures to tackle the challenges of violent extremism. As President Xi Jinping has pointed out, 'the realization of the full enjoyment of human rights by human beings is the common goal of human society.' While establishing a community of common destiny for all mankind, the international community should take action to effectively address the root causes of violent extremism, maintain a peaceful and secure development environment, and defend the basic human rights and well-being of all peoples, so as to ensure a safe and brighter future for all.

Facts Show July 5 Riot “Far Beyond Ordinary Violent Matter”

Nearly two weeks after the July 5 riot in Urumqi of northwest China＇s Xinjiang Uygur Autonomous Region, social order and people＇s lives are returning to normal. Yet in retrospect, a mass of evidences show that the unrest was a wellplanned violent criminal incident of terrorist nature.

Violent Times

THE news of the fatal Valentine＇s Day shooting in South Africa of celebrity Reeva Steenkamp by her boyfriend, the world athletic icon Oscar Pistorius, has shocked the country to its core. That shock was profound because Pistorius is the personifica- tion of the spirit of determination over adversity, and is someone who has transcended his physical disability to become a global sports champion and role model for all South Africans.

Determination of HCH and DDT in soils following Wenchuan 5.12 Violent Earthquake in China

A large number of disinfectants and insecticides were used on a large scale after the Wenchuan 5.12 Violent Earthquake. The disinfectants and insecticides, including HCH (hexachlorocyclohexane) and DDT (dichlorodiphenyltrichoroethane) as representative OCPs, could lead to serious environmental problems. In order to disclose the residues of HCH and DDT in soils following the Wenchuan 5.12 Violent Earthquake, surface soils in Dujiangyan City and Beichuan County were sampled in 2009. The samples were analyzed by GC/MS after pre- treatment by ultrasonic solvent extraction. It has been found that the concentra-tions of γ-HCH, β-HCH and 4,4'-DDT in soils of Beichuan County were 94–102.4, 14.1–36.1 and 2.9–4.3 times high their concentrations in those of the Wolong Nature Reserve, and the concentrations in soils of Dujiangyan City were 52.4–61.6, 8.8 –45.3 and 5.4–6.2 times high their concentrations in soils of the Wolong Nature Reserve, respectively. According to the isomers of HCHs and metabolites of DDTs, it could be concluded that these areas may have additional inputs of lindane and 4,4'-DDT.

基于Transformer模型的“暴力”虚开发票风险识别

自2016年“营改增”全面实施以来,与之相关的免税减税等税收优惠政策原旨在惠企助企、激发市场活力,但不法分子在巨额利润驱动下企图通过虚开增值税发票骗取出口退税、抵扣税款,严重扰乱了税收秩序。本文以“暴力”虚开发票的企业的犯罪特征为切入点,从基础征管数据和增值税发票数据中选取了24项虚开指标,构建了基于Transformer模型的虚开增值税发票识别模型,对虚开公司进行检测。实证分析表明Transformer模型对虚开增值税发票的识别召回率为0.934 7,准确率为0.986 9,AUC为0.963 9,显著优于SVM、Xgboost、MLP等传统机器学习模型,可辅助税务部门高效识别“暴力”虚开企业,节省人工筛查成本,对有效打击虚开增值税发票一类违法犯罪行为具有非常重要的实践意义。

基于对比学习的暴力检测算法研究

针对现有暴力检测模型实际部署中存在数据标注成本高,提出基于对比学习的半监督模型训练框架,利用对比学习训练模型的表征能力,对比样本采用基于速度和基于全局和局部对比生成方式,对比框架在大量正负样本对比基础上增加正例样本间对比数量,同时利用伪标注对模型进行微调。实验结果显示,对比学习能够帮助模型在RWF2000和RVLS 5%训练数据下提升了3.9%,2.55%准确率,微调阶段能在RWF200025%训练数据下帮助模型进一步提升约3%准确率。

简·亚当斯和平主义思想溯源及实践回顾

本文追溯了简·亚当斯和平主义思想生发的现实背景和精神源头,包括个性传承、宗教伦理观以及依据良心而活的信念。长于观察、阅读、反思的亚当斯在生活和书本的双重熏陶中获得了洞见和力量,冲破家庭束缚和社会成规,创建赫尔会馆,参加进步运动,反对美国政府发动的美西、美菲战争,并在一战期间积极维护和平,即便遭遇声名沉浮也坚守和平主义理念,在她所处的时代展现了女性独立意识以及拒绝暴力与战争的人道主义精神。

从铁笼囚禁到文化植入——论《致某科学院的报告》中的权力规训

卡夫卡《致某科学院的报告》是一份受虐狂向施虐狂汇报受虐体验的报告,其背后隐藏着一座由技术官僚构成、假科学名义施暴的权力城堡。猿性而人名的报告者红彼得只是权力生产的结果,权力才是该小说隐身不现的主角。最高官僚权力机构科学院通过各级暴力运作对红彼得实施规训与操控,强迫其就范,诱惑其为奴;红彼得在报告中所炫耀的进步与成功,究其实质不过是它在权力狡计之下从受虐转变为自虐的身心自戕而已。官僚机构的权力规训由机械打击的硬暴力与文化植入的软暴力相配合完成,其规训效果则表现为红彼得患上斯德哥尔摩综合征,迎合与膜拜权力并自我驯化。

Preventing and Countering Violent Extremism Through Education:A Vital Element for the Peace Process in the Gulf

While the fall of Raqqa signalled the steady decline and“defeat”of the Islamic State(IS),it did not guarantee the end of its ideology of violent extremism.The possibility of the organisation reinventing itself in different and perhaps more lethal ways is very high.States,therefore,must adopt measures that are based on holistic,multi-stakeholder approaches rather than on military and securitised responses to counter and ultimately prevent the spread of violent extremist ideology.The countries of the Gulf have the advantage of a“youth bulge,”with almost half its population under the age of 30 years.This“youth bulge”is however a double-edged sword.This large pool of human capital could be a demographic dividend,if harnessed intelligently,while conversely it could prove to be a demographic disaster if timely and plausible solutions to the various issues faced are not provided.This paper will explore ways in which preventing and countering violent extremism(P/CVE)through education could be achieved by including a variety of stakeholders from faith based organisations,the corporate sector,media and academia in the formulation of appropriate educational responses which take into account issues like culture,identity and religiosity.These re-education programmes should form an integral part of P/CVE policies,thereby providing the much needed psycho-social angle to an otherwise hard security-driven discourse.It is against this rationale that this paper makes a case for P/CVE through education,encompassing the formal,informal and non-formal levels,as one of the possible means for promoting and fostering a peace process in the Gulf.

网络暴力言论检测的技术和实践

暴力言论是网络语言治理的重点,由于社交媒体信息量巨大,运用自然语言处理等人工智能技术对网络进行自动检测是及时阻止网络暴力言论传播扩散的必然要求。由于网络暴力言论表达具有隐匿性和多样性,目前存在缺乏细粒度的中文暴力言论语料库,以及对于采取隐式表达的暴力言论检测效果不理想的问题。因此,本文构建了一个大规模的中文暴力言论语料库,并研制了基于双重对比学习和基于情绪辅助的多标签自训练暴力言论检测方法,并在多种数据集上验证了其有效性和先进性。在大语言模型时代,将其与背景知识相结合可以更准确地检测暴力言论,但其生成内容需要与人类价值观对齐以符合当今的社会道德标准。暴力言论检测研究在未来发展包括多模态暴力言论检测研究、基于大模型的暴力言论检测方法研究、基于大模型的暴力言论监管机制研究、暴力言论检测的可解释性研究等4个主要方向。

基于改进分层注意网络和TextCNN联合建模的暴力犯罪分级算法

为了科学、智能地对服刑人员的暴力倾向分级,将自然语言处理(NLP)中的文本分类方法引入犯罪心理学领域,提出一种基于改进分层注意网络(HAN)与TextCNN(Text Convolutional Neural Network)两通道联合建模的犯罪语义卷积分层注意网络(CCHA-Net),通过分别挖掘犯罪事实与服刑人员基本情况的语义信息,完成暴力犯罪气质分级。首先,采用Focal Loss同时替代两通道中的Cross-Entropy函数,优化样本数量不均衡问题。其次,在两通道输入层中,同时引入位置编码,改进对位置信息的感知能力;改进HAN通道,采用最大池化构建显著向量。最后,输出层都采用全局平均池化替代全连接方法,以避免过拟合。实验结果表明,与AC-BiLSTM(Attention-based Bidirectional Long Short-Term Memory with Convolution layer)、支持向量机(SVM)等17种相关基线模型相比,CCHA-Net各项指标均最优,微平均F1(Micro_F1)为99.57%,宏平均和微平均下的曲线下面积(AUC)分别为99.45%和99.89%,相较于次优的AC-BiLSTM提高了4.08、5.59和0.74个百分点,验证了CCHA-Net能有效胜任暴力犯罪气质分级任务。

骨科术后患儿疼痛灾难化现状及其影响因素分析

目的探讨骨科术后患儿疼痛灾难化现状及其影响因素。方法选取2023年5月-10月湖南省3所三级甲等医院的骨科术后患儿321例为研究对象,采用一般资料调查表、疼痛灾难化量表、视觉模拟评分量表、恐动症量表及骨科患者功能锻炼依从性量表进行调查。结果骨科术后患儿疼痛灾难化总分为(38.28±2.47)分,处于较高水平。多元线性回归分析结果显示,骨科术后患儿疼痛灾难化的影响因素包括手术方式、术后天数、疼痛、恐动症及功能锻炼依从性(P<0.05)。结论骨科术后患儿疼痛灾难化处于较高水平,医护人员应根据上述影响因素重点关注骨科术后患儿疼痛灾难化水平,制定有效的干预方案,促进患儿术后康复。

面向筛面复杂背景的矿山异物视觉检测方法

矿山异物检测是异物智能化去除的前提,更是保障设备安全运行,矿山正常生产的关键。在矿山生产过程中,异物来源广泛,种类繁杂。针对传统的矿山异物检测方法面临适应性差和效率低的问题,提出了一种面向大型振动筛筛面的矿山异物检测算法模型。为解决强振动、矿石遮挡和粉尘水雾等复杂环境的干扰,该模型引入了改进的显式视觉中心模块(EVCBlock),轻量化上采样算子CARAFE和基于动态非单调聚焦机制的梯度增益损失函数WiseIoU-v3,有效提升了在复杂环境下的异物检测性能。利用TensorRT对模型优化并部署至边缘计算设备Jetson Xavier NX,实现了在边缘侧的异物检测。研究结果表明:该模型在振动筛筛面异物检测上的表现明显好于其他对比模型。经多线程视频推流测试,模型部署至边缘计算设备平均识别精确率可以达到96.3%,平均帧率达到25 FPS以上,满足了实际检测要求。

An unstructured mesh method for numerical simulation of violent sloshing flows

An unstructured mesh Reynolds-averaged Navier-Stokes(RANS)solver has been developed for numerical simulation of violent sloshing flows inside a tank with complicated inner structures.The numerical solver employs the unstructured multi-dimensional tangent hyperbolic interface capturing method(UMTHINC)for free-surface capturing combined with various turbulence models.The sloshing motion is numerically modeled using the body-force method which introduces a source term into the momentum equation corresponding to the tank motion profile.Numerical simulations of the tank sloshing problems are performed for different test cases with various oscillation frequencies.The performance of the interface capturing method has been discussed and the effect of turbulence model choice on loading predictions is highlighted by studying several RANS models and analyzing its effect on fluid motion and impact pressure.Numerical simulations of the sloshing inside the tank with a vertical baffle has also been conducted and a discussion is provided on different numerical treatment of the baffle.