Linguistic Patterns for Code Word Resilient Hate Speech Identification

The permanent transition to online activity has brought with it a surge in hate speech discourse. This has prompted increased calls for automatic detection methods, most of which currently rely on a dictionary of hate speech words, and supervised classification. This approach often falls short when dealing with newer words and phrases produced by online extremist communities. These code words are used with the aim of evading automatic detection by systems. Code words are frequently used and have benign meanings in regular discourse, for instance, “skypes, googles, bing, yahoos” are all examples of words that have a hidden hate speech meaning. Such overlap presents a challenge to the traditional keyword approach of collecting data that is specific to hate speech. In this work, we first introduced a word embedding model that learns the hidden hate speech meaning of words. With this insight on code words, we developed a classifier that leverages linguistic patterns to reduce the impact of individual words. The proposed method was evaluated across three different datasets to test its generalizability. The empirical results show that the linguistic patterns approach outperforms the baselines and enables further analysis on hate speech expressions.

Download Full-text

Countering hate speech through arts and arts education

Policy Futures in Education ◽

10.1177/1478210319848953 ◽

2019 ◽

Vol 18 (3) ◽

pp. 344-357 ◽

Cited By ~ 1

Author(s):

Tuula Jääskeläinen

Keyword(s):

Arts Education ◽

Hate Speech ◽

Freedom Of Expression ◽

Legal Framework ◽

Global Scale ◽

Online Activity ◽

Online Systems ◽

Counter Narratives ◽

The Arts ◽

The Impact

Hate speech has become a growing topic of discussion and debate on a global scale, especially as advances in the internet transform communication on many levels. Among scholars, hate speech has been defined as any form of expression – for example by means of speech, images, videos or online activity – that has the capacity to increase hatred against a person or people because of a characteristic they share or a group to which they belong. In order to maintain the integrity of a functioning democracy, it is important to identify the best balance between allowing freedom of expression and protecting other human rights by countering hate speech. In addition to strengthening the legal framework to address the cases when hate speech can be considered criminal, and developing automated monitoring of online systems to prevent the spreading of cyberhate, counter narratives can be utilized by the targets of hate speech and their communities to create campaigns against hate speech. The employment of artists’ expression and arts education have great potential for creating different counter narratives to challenge one-sided narratives and hate speakers’ simplified generalizations. Because hate speech is not an easy issue to address in schools, clear research evidence, concrete guidelines and practical examples can help teachers to contribute, along with their students, in combating it. A great body of evidence supporting the beneficial social impacts of the arts and culture fields is already available, but much more research, backed by sufficient resources, is needed to evaluate the impact and effectiveness of intervention strategies in countering hate speech through arts education.

Download Full-text

The Effects of Employment Protection legislationon Personnel Policies: A Review of Theoretical Modelsand Empirical results

Voprosy Ekonomiki ◽

10.32609/0042-8736-2014-3-126-140 ◽

2014 ◽

pp. 126-140

Author(s):

O. Mironenko

Keyword(s):

Personnel Selection ◽

Selection Criteria ◽

Theoretical Models ◽

Employment Protection ◽

Empirical Results ◽

Protection Legislation ◽

Labour Contracts ◽

Hiring And Firing ◽

The Impact ◽

Personnel Policies

Employers incur costs while fulfilling the requirements of employment protection legislation. The article contains a review of the core theoretical models and empirical results concerning the impact of these costs on firms’ practices in hiring, firing, training and remuneration. Overall, if wages are flexible or enforcement is weak, employment protection does not significantly influence employers’ behavior. Otherwise, stringent employment protection results in the reduction of hiring and firing rates, changes in personnel selection criteria, types of labour contracts and dismissal procedures, and, in some cases, it may lead to the growth of wages and firms’ investments to human capital.

Download Full-text

Effectiveness of Supervised Classification Models for Hate Speech on Twitter

10.36227/techrxiv.13140281 ◽

2020 ◽

Author(s):

Kunal Srivastava ◽

Ryan Tabrizi ◽

Ayaan Rahim ◽

Lauryn Nakamitsu

Keyword(s):

Supervised Classification ◽

Hate Speech ◽

The Internet ◽

Classification Models ◽

Personal Attack ◽

Political Beliefs ◽

The Many ◽

Primary Medium ◽

Offensive Speech

<div> <div> <div> <p>Abstract </p> <p>The ceaseless connectivity imposed by the internet has made many vulnerable to offensive comments, be it their physical appearance, political beliefs, or religion. Some define hate speech as any kind of personal attack on one’s identity or beliefs. Of the many sites that grant the ability to spread such offensive speech, Twitter has arguably become the primary medium for individuals and groups to spread these hurtful comments. Such comments typically fail to be detected by Twitter’s anti-hate system and can linger online for hours before finally being taken down. Through sentiment analysis, this algorithm is able to distinguish hate speech effectively through the classification of sentiment. </p> </div> </div> </div>

Download Full-text

Un-Supervised approach to backorder prediction using deep autoencoder

Recent Patents on Computer Science ◽

10.2174/2213275912666190819112609 ◽

2019 ◽

Vol 12 ◽

Author(s):

Gunjan Saraogi ◽

Deepa Gupta ◽

Lavanya Sharma ◽

Ajay Rana

Keyword(s):

Anomaly Detection ◽

Predictive Model ◽

Inventory Management ◽

Supervised Classification ◽

Operating Characteristic ◽

Imbalanced Data ◽

Bullwhip Effect ◽

Detection Methods ◽

Good Representation ◽

Operational Management

Background: Backorders are an accepted abnormality affecting accumulation alternation and logistics, sales, chump service, and manufacturing, which generally leads to low sales and low chump satisfaction. A predictive archetypal can analyse which articles are best acceptable to acquaintance backorders giving the alignment advice and time to adjust, thereby demography accomplishes to aerate their profit. Objective: To address the issue of predicting backorders, this paper has proposed an un-supervised approach to backorder prediction using Deep Autoencoder. Method: In this paper, artificial intelligence paradigms are researched in order to introduce a predictive model for the present unbalanced data issues, where the number of products going on backorder is rare. Result: Un-supervised anomaly detection using deep auto encoders has shown better Area under the Receiver Operating Characteristic and precision-recall curves than supervised classification techniques employed with resampling techniques for imbalanced data problems. Conclusion: We demonstrated that Un-supervised anomaly detection methods specifically deep auto-encoders can be used to learn a good representation of the data. The method can be used as predictive model for inventory management and help to reduce bullwhip effect, raise customer satisfaction as well as improve operational management in the organization. This technology is expected to create the sentient supply chain of the future – able to feel, perceive and react to situations at an extraordinarily granular level

Download Full-text

The Effects of Megaproject Social Responsibility on Participating Organizations

Project Management Journal ◽

10.1177/87569728211015868 ◽

2021 ◽

pp. 875697282110158

Author(s):

Hanyang Ma ◽

Daxin Sun ◽

Saixing Zeng ◽

Han Lin ◽

Jonathan J. Shi

Keyword(s):

Social Responsibility ◽

Social Performance ◽

Empirical Results ◽

Positive Effects ◽

The Impact

This study focuses on the effects of megaproject social responsibility (MSR) on participating organizations’ performance. Using a survey of the participating organizations in Chinese megaprojects, this study reveals that the impact of MSR on a participant’s performance goes beyond the scope of the current megaproject. The empirical results indicate that MSR positively affects both financial and social performance of the participating organizations. The interactions of primary stakeholders weaken the positive effects of MSR on both financial and social performance, whereas the interactions of secondary stakeholders strengthen the positive effects of MSR on social performance.

Download Full-text

A Systematic Review of Landsat Data for Change Detection Applications: 50 Years of Monitoring the Earth

Remote Sensing ◽

10.3390/rs13152869 ◽

2021 ◽

Vol 13 (15) ◽

pp. 2869

Author(s):

MohammadAli Hemati ◽

Mahdi Hasanlou ◽

Masoud Mahdianpari ◽

Fariba Mohammadimanesh

Keyword(s):

Time Series ◽

Change Detection ◽

Policy Development ◽

Atmospheric Correction ◽

Open Data ◽

Detection Methods ◽

Landsat 8 ◽

Landsat Images ◽

Detection Analysis ◽

The Impact

With uninterrupted space-based data collection since 1972, Landsat plays a key role in systematic monitoring of the Earth’s surface, enabled by an extensive and free, radiometrically consistent, global archive of imagery. Governments and international organizations rely on Landsat time series for monitoring and deriving a systematic understanding of the dynamics of the Earth’s surface at a spatial scale relevant to management, scientific inquiry, and policy development. In this study, we identify trends in Landsat-informed change detection studies by surveying 50 years of published applications, processing, and change detection methods. Specifically, a representative database was created resulting in 490 relevant journal articles derived from the Web of Science and Scopus. From these articles, we provide a review of recent developments, opportunities, and trends in Landsat change detection studies. The impact of the Landsat free and open data policy in 2008 is evident in the literature as a turning point in the number and nature of change detection studies. Based upon the search terms used and articles included, average number of Landsat images used in studies increased from 10 images before 2008 to 100,000 images in 2020. The 2008 opening of the Landsat archive resulted in a marked increase in the number of images used per study, typically providing the basis for the other trends in evidence. These key trends include an increase in automated processing, use of analysis-ready data (especially those with atmospheric correction), and use of cloud computing platforms, all over increasing large areas. The nature of change methods has evolved from representative bi-temporal pairs to time series of images capturing dynamics and trends, capable of revealing both gradual and abrupt changes. The result also revealed a greater use of nonparametric classifiers for Landsat change detection analysis. Landsat-9, to be launched in September 2021, in combination with the continued operation of Landsat-8 and integration with Sentinel-2, enhances opportunities for improved monitoring of change over increasingly larger areas with greater intra- and interannual frequency.

Download Full-text

Screening for Melanoma and Other Skin Cancer Shows a Higher Early Melanoma Incidence: Social Educational Program “Life Fear-Free”

Dermatopathology ◽

10.3390/dermatopathology8010011 ◽

2021 ◽

Vol 8 (1) ◽

pp. 54-68

Author(s):

Lev Demidov ◽

Igor Samoylenko ◽

Nina Vand ◽

Igor Utyashev ◽

Irina Shubina ◽

...

Keyword(s):

Skin Cancer ◽

Screening Program ◽

Excisional Biopsy ◽

Skin Lesions ◽

Detection Methods ◽

Non Melanoma Skin Cancer ◽

The Social ◽

Educational Campaign ◽

Skin Cancer Detection ◽

The Impact

Background: The screening program Life Fear-Free (LFF) aimed at early diagnosis of cutaneous melanoma (CM) was introduced in Samara, Chelyabinsk, Yekaterinburg, and Krasnodar (Russia) in 2019. Objectives: To analyze the impact of the program on early CM and non-melanoma skin cancer (NMSC) detection. Methods: According to the social educational campaign, people were informed about CM risk factors and symptoms and were invited for skin examination. The program planned to involve 3200 participants in total. Participants with suspicious lesions were invited for excisional biopsy. Results: 3143 participants, including 75.4% women, were examined for skin lesions. The average age of the participants was 43.7 years. Mostly skin phototypes II and III were registered (48.2% and 41.0%, respectively); 3 patients had CM, 15 had basal cell carcinoma, and 1 had Bowen’s disease, which were confirmed histologically. All detected melanomas had Breslow’s thickness of 1 mm. Conclusion: The participants showed high interest in early skin cancer detection programs. The incidence rate of CM and NMSCs among the program participants was higher than in general public. The early disease grade was proven for the detected CMs and NMSCs. The study has shown that it is important to continue such programs.

Download Full-text

Emotionally Informed Hate Speech Detection: A Multi-target Perspective

Cognitive Computation ◽

10.1007/s12559-021-09862-5 ◽

2021 ◽

Author(s):

Patricia Chiril ◽

Endang Wahyu Pamungkas ◽

Farah Benamara ◽

Véronique Moriceau ◽

Viviana Patti

Keyword(s):

Hate Speech ◽

Binary Classification ◽

Online Communication ◽

Vulnerable Groups ◽

Speech Detection ◽

Social Media Platforms ◽

Sentic Computing ◽

Specific Manifestation ◽

The Impact ◽

First Time

AbstractHate Speech and harassment are widespread in online communication, due to users' freedom and anonymity and the lack of regulation provided by social media platforms. Hate speech is topically focused (misogyny, sexism, racism, xenophobia, homophobia, etc.), and each specific manifestation of hate speech targets different vulnerable groups based on characteristics such as gender (misogyny, sexism), ethnicity, race, religion (xenophobia, racism, Islamophobia), sexual orientation (homophobia), and so on. Most automatic hate speech detection approaches cast the problem into a binary classification task without addressing either the topical focus or the target-oriented nature of hate speech. In this paper, we propose to tackle, for the first time, hate speech detection from a multi-target perspective. We leverage manually annotated datasets, to investigate the problem of transferring knowledge from different datasets with different topical focuses and targets. Our contribution is threefold: (1) we explore the ability of hate speech detection models to capture common properties from topic-generic datasets and transfer this knowledge to recognize specific manifestations of hate speech; (2) we experiment with the development of models to detect both topics (racism, xenophobia, sexism, misogyny) and hate speech targets, going beyond standard binary classification, to investigate how to detect hate speech at a finer level of granularity and how to transfer knowledge across different topics and targets; and (3) we study the impact of affective knowledge encoded in sentic computing resources (SenticNet, EmoSenticNet) and in semantically structured hate lexicons (HurtLex) in determining specific manifestations of hate speech. We experimented with different neural models including multitask approaches. Our study shows that: (1) training a model on a combination of several (training sets from several) topic-specific datasets is more effective than training a model on a topic-generic dataset; (2) the multi-task approach outperforms a single-task model when detecting both the hatefulness of a tweet and its topical focus in the context of a multi-label classification approach; and (3) the models incorporating EmoSenticNet emotions, the first level emotions of SenticNet, a blend of SenticNet and EmoSenticNet emotions or affective features based on Hurtlex, obtained the best results. Our results demonstrate that multi-target hate speech detection from existing datasets is feasible, which is a first step towards hate speech detection for a specific topic/target when dedicated annotated data are missing. Moreover, we prove that domain-independent affective knowledge, injected into our models, helps finer-grained hate speech detection.

Download Full-text

A noise reduction method for semi-supervised community detection based on harmonic function

International Journal of Modern Physics B ◽

10.1142/s0217979218501667 ◽

2018 ◽

Vol 32 (14) ◽

pp. 1850166 ◽

Cited By ~ 1

Author(s):

Lilin Fan ◽

Kaiyuan Song ◽

Dong Liu

Keyword(s):

Harmonic Function ◽

Prior Knowledge ◽

Community Detection ◽

Detection Methods ◽

Detection Process ◽

Knowledge Based ◽

Detection Approach ◽

Series Of Experiments ◽

Novel Strategy ◽

The Impact

Semi-supervised community detection is an important research topic in the field of complex network, which incorporates prior knowledge and topology to guide the community detection process. However, most of the previous work ignores the impact of the noise from prior knowledge during the community detection process. This paper proposes a novel strategy to identify and remove the noise from prior knowledge based on harmonic function, so as to make use of prior knowledge more efficiently. Finally, this strategy is applied to three state-of-the-art semi-supervised community detection methods. A series of experiments on both real and artificial networks demonstrate that the accuracy of semi-supervised community detection approach can be further improved.

Download Full-text

Automatic Handgun Detection with Deep Learning in Video Surveillance Images

Applied Sciences ◽

10.3390/app11136085 ◽

2021 ◽

Vol 11 (13) ◽

pp. 6085

Author(s):

Jesus Salido ◽

Vanesa Lomas ◽

Jesus Ruiz-Santaquiteria ◽

Oscar Deniz

Keyword(s):

Neural Network ◽

Deep Learning ◽

Convolutional Neural Network ◽

Video Surveillance ◽

Automatic Detection ◽

Public Spaces ◽

Detection Methods ◽

Training Dataset ◽

Average Precision ◽

Terrorist Acts

There is a great need to implement preventive mechanisms against shootings and terrorist acts in public spaces with a large influx of people. While surveillance cameras have become common, the need for monitoring 24/7 and real-time response requires automatic detection methods. This paper presents a study based on three convolutional neural network (CNN) models applied to the automatic detection of handguns in video surveillance images. It aims to investigate the reduction of false positives by including pose information associated with the way the handguns are held in the images belonging to the training dataset. The results highlighted the best average precision (96.36%) and recall (97.23%) obtained by RetinaNet fine-tuned with the unfrozen ResNet-50 backbone and the best precision (96.23%) and F1 score values (93.36%) obtained by YOLOv3 when it was trained on the dataset including pose information. This last architecture was the only one that showed a consistent improvement—around 2%—when pose information was expressly considered during training.

Download Full-text