Hate Speech on Social Media in Malaysia - A Thematic Review

Vol.


Introduction
Malaysia is undeniably one of the countries striving for technological modernity in this age of information and technology. Today's technology is evolving in terms of sophistication. Traditional details give way over time to more modern technical information, in line with the flow of modernity. One of the latest technologies are social networks that allow us to socialize and communicate with people from all over the world.
According to Petterson (2020), malicious postings or hate speech have often been investigated in the context of communications by populist (far-right) leaders. This statement is supported by Kumar et al (2017), who said that these gamers are also increasingly associated with the uncontrolled spread of misinformation or fake news. As it is used by millions of people or netizens, some negative things like spreading false information, identity fraud and most importantly hate speech will get out of control because social media platforms like twitter never have a guardian. Since these platforms partially guarantee users' freedom of expression, they can be used to attack or insult others. Kumar et al (2017) noted that online attacks and abuse have been shown to cause victims not only psychological and physiological health problems, but also self-harm and suicide. Therefore, our approach is to analyze hate speech intent in controlling hate speech in cyber world to prevent further negative impact on users.
According to Brown (2017), hate speech or malicious language is used to identify messages that violate existing legal norms and require government regulation or a publicly shared statement that incites, encourages, justifies, or justifies hate, discrimination or hostility towards a particular group such as race or tribal, belief, ability or disability, gender, age, sexual orientation or gender identity. Apart from that, Kilvington (2020) defined that hate speech is defined as disgusting when it spreads, incites and promotes hatred, violence and discrimination against individuals or groups because of their protected status. These include race, ethnicity, religion, gender, sexual orientation, and disability. Therefore, the objective of this paper is to delve of discussion of hate speech on social media that been discussed in the hate speech publications from the year 2018-2022 as through the following research question RQ: What are the discussion of hate speech on social media found in the literature from 2018-2022 in Malaysia?

Materials and Methods
The term thematic review using ATLAS.ti 22 as the tool as being introduced by Zairul (2020) is implemented because the method of this study applies thematic analysis procedure in a literature review. Clarke & Braun (2013) define thematic analysis is a process of identifying the pattern and construct themes over thorough reading on the subject. The following step is to identify the pattern and construct category to understand the trend of IBS publication in the country. The tenets of the research are to analyse and interpret the findings for the recommendation of future research in hate speech subject. The selection of literature was performed according to several selection criteria: 1) publication from 2018-2022, 2) Have at least keyword(s) hate speech AND Malaysia, 3) Focusing on hate speech issues in Malaysia. The decision to limit the country of origin was made to help define the issues and problems of hate speech in the Malaysia context. Boolean operator: Use OR, AND or NOT. "hate " (All Fields) AND "speech" (All Fields) AND "Malaysia" (All Fields) 25 results

papers used
The literature search was performed in the Scopus, Research Gate and WoS search. The initial search came out with 8 articles from SCOPUS, 20 articles from Reseacrh Gate and 25 articles from WoS. However, 33 articles were removed due to their premature results and anecdotes or were not discussing hate speech issues in Malaysia. Some of the articles were also found incomplete, or the full articles are not accessible, have a broken link and overlapped and metadata incomplete Therefore, the final paper to be reviewed down to 23 articles as said in table 1.0.
The articles were uploaded in the ATLAS.ti 23 as primary documents, and then each paper was grouped into 1) author; 2) issue number; 3) periodical, 4) publisher, 5) volume and 6) year of publication. In doing so, the articles can be analysed according to the year it was published and what is the discussion pattern according to the year. The total articles finalised into the final documents in the ATLAS.ti 23 is 20 documents. Only English paper were selected, and to make it easier to understand, researcher provide figure below.

1) Results and Discussions
The main conclusions of the thematic review are reported in this section.   To answer the RQ above, seven out of 20 papers are in the theme of hate speech detection which is 35% of the data that researchers revealed. 55% of the data, on the other hand, researcher put in the theme of hate speech issues and concerns, with eleven out of twenty literatures. Meanwhile, for the hate speech in Covid-19 pandemic, 10% of the data in included in the theme and only two of the papers stated this theme.This trend make researcher want to explore more about hate speech on social media especially in Malaysia.
A lot researcher (Teh & Cheng, 2020;Mullah & Zainon, 2021;Bilal et al., 2022) that come from data analytic or information technologies background want to explore more on tracing and tracking the hate speech itself. Different with some scholars (Saufi et al., 2022;Nor & Gale, 2021;Nordin, 2022) from communication and psychology backgrounds, they want to delve into the issues and concerns behind hate speech. There are various angle they can study such as LGBT issue, Islamisation and religious value. For the third theme, hate speech in Covid-19 pandemic, researchers (Zamri et al., 2021;Fernandez, 2020) come various kind of background like science politic and also information technologies. For the topic, researcher are a bit restricted towards the timeline because they only allowed to do research only from earlier 2020 until earlier 2022.

Theme 1: Hate Speech Detection
The literature from 2018-2022 in Malaysia discusses the development and effectiveness of hate speech detection algorithms on social media platforms, such as Twitter and Facebook. Various approaches have been proposed, including machine learning techniques and rulebased methods, to identify hate speech content and reduce its spread.
Hate speech detection refers to the process of identifying and flagging instances of hate speech in written or spoken language. It involves using natural language processing and machine learning techniques to analyze and classify text or speech as containing hate speech or not. The aim of hate speech detection is to prevent the spread of hateful or harmful content on social media platforms or other online spaces by identifying and removing such content. Hate speech detection is an important tool for maintaining a safe and respectful online environment that promotes free expression without causing harm or promoting discrimination.
In accordance to Teh & Cheng (2020), profanity is more associated with hatred, profanity-based methods in hate speech detection could be effective and profanity screening can address the class imbalance issue in hate speech detection. Allign with that Ibrahim et al (2021) present an image in a series of text describing the top words. This model can be considered as a simple way to exchange high-level information without overloading the user's details in expressing hate speech on social media.
For Mullah & Zainon (2021), they make a quite big contributions, as they come out with three-fold of results. First is to equip the readers with the necessary information on the critical steps involved in hate speech detection using machine learning (ML) algorithms. Secondly, the weaknesses and strengths of each method is critically evaluated to guide researchers in the algorithm choice dilemma. Lastly, some research gaps and open challenges were identified. The different variants of ML techniques were reviewed which include classical ML, ensemble approach and deep learning methods. Mansur et al. (2023) focused on presenting an in-depth understanding of current perspectives and highlighted research opportunities to boost the quality of hate speech detection systems.
In other study that conducted by Aziz et al (2021)  According to Mullah & Zainon (2022) in their second paper that have been reviewed, the researchers built a novel heterogeneous stacked ensemble (HSE) classifier for detecting politically motivated cyber-hate on Twitter. We constructed a heterogeneous stacked ensemble with eight baseline estimators. Hence, all of this literature that conduct have highly recommended tools or ways in track and trace all kind of hate speech words on social media. Researchers of the papers also provide huge contribution to the research of hate speech as whole.

Hate Speech Issues and Concerns
The literature also highlights the prevalence of hate speech on social media in Malaysia, which has been linked to rising levels of religious and racial tension. Scholars argue that hate speech online can fuel discrimination, prejudice, and even violence, and call for increased efforts to combat this problem.
Hate speech refers to any speech or expression that attacks, threatens, or insults a person or a group of people based on their race, ethnicity, nationality, religion, gender, sexual orientation, or any other characteristic that defines them. Hate speech issues and concerns arise when such speech is used to promote intolerance, discrimination, or violence against a targeted group or individual. The impact of hate speech can be far-reaching and harmful, leading to social exclusion, harassment, and even physical harm. It can also have negative consequences for individuals' mental health and well-being. Moreover, hate speech can undermine social cohesion and fuel social tensions, leading to broader societal conflicts.
The first scholar, Saufi et al (2022) find out that the use of social media platforms is on the rise, increasing the possibility to infringe on fundamental liberties such as hate speech and expression. Apart from that, Wan Mohd Nor and Peter Gale (2021) in their study show that Islamisation tends to be framed negatively, contributing to the fear of Islamisation and restriction of fundamental freedoms. Allign with that, there are a number of religious values related to courtesy and morality in the concept, which pertain to the prevention of hate speech. The findings also show that the seven enriching values under the characteristic of inclusivity in the concept, need effective strategies for preventing hate speech, such as enhancing the role of religious leaders and the media (Nordin, 2022).
In other study that conducted by same scholar, Nor (2018) concludes that by putting forth a holistic approach to the application of free speech that may be best to applied in applied a multi-cultural and multi-religious country. It is hoped that by emphasizing on the need to exercise free speech with restraint and responsibility, the people will slowly come to appreciate the values of propriety and basic human decency. Mia et al (2022) emphasizes the importance of information in various platforms such as social media for citizens and the limitations of their freedom of speech and expressions. Azman and Zamri (2022) in their study stated that both Dramaturgical Models of social interaction (Goffman, 1959) and Cultivation Theory (Gerbner, 1973) were used to explain the hate speech culture phenomenon especially in social media. Meanwhile, Sharma (2020) argues that although several legal provisions are available to curb hate speech, Alternative Dispute Resolution and counter speech may be used as alternative ways to address hate speech.
In the same year, Lingam et al (2019) concludes that most of the flamers, that refer to hate speech utterers, kept their identity anonymous due privacy concerns and for the freedom of speech. The rest of the flamers used their real name as a form of publicity, identity defining and to boost their self-confidence. Irony with the previous research, Thun et al (2022) proposes a mechanism, which combines the best cyberbullying detection features to fill the gaps and limitations of existing applications. The results of the study have shown that the proposed mobile application records a higher accuracy in detecting cyberbully than other available applications. Ismail et al (2020) offers some novel insights into the field of impoliteness by drawing attention to online users' defensive linguistic behaviour in dealing with out-group impoliteness strategies in online communication. Zawawi et al (2021) study show how the interpretations and reactions of Malaysian audiences on LGBT issues in turn provide a preliminary picture of how the effects of media exposure indirectly influence the psychology of individuals.
Therefore, hate speech issues and concerns are particularly acute in the online context, where anonymity and the ease of sharing content can contribute to the proliferation of hateful content. Hate speech online can cause harm to individuals and groups and create a toxic environment that undermines civil discourse and democratic values. Addressing hate speech issues and concerns requires a multifaceted approach, involving education, awareness-raising, regulation, and community engagement to promote tolerance, respect, and inclusivity.

Hate Speech in Covid-19 Pandemic Era
In the context of the Covid-19 pandemic, the literature discusses the emergence of hate speech related to the virus and its impact on marginalized communities in Malaysia. There have been reports of discriminatory and racist messages targeting specific groups, such as migrant workers and minorities, on social media platforms. Scholars emphasize the need for greater awareness and education to promote tolerance and understanding during this difficult time.
Hate speech in the COVID-19 pandemic refers to the spread of discriminatory or prejudicial language related to the pandemic and its impact on individuals and communities. It includes speech that targets specific groups, such as people of certain nationalities or ethnicities, blaming them for the outbreak or spreading misinformation about them. Hate speech related to the COVID-19 pandemic has been fueled by fear, misinformation, and social and economic tensions. It has led to discrimination and stigmatization against marginalized communities, including migrants, refugees, and minorities. This has created additional challenges for these groups, including limited access to healthcare, education, and social services.
For the literature, Zamri et al (2021) actually contributes to the academic debates about the public sphere and social media's role in constructing meaning in cultural and social change. It also suggests that Twitter develops policies to prohibit hate speech and impose regulations to ensure that online spaces remain civil, safe, and democratic. Other than that Fernandez (2020) said there is an abstract corelation between populist exclusionary statements by political leaders and the responding hate speech sentiments being shared online during the pandemic.
Hence, The impact of hate speech during the COVID-19 pandemic can be significant and harmful, leading to social exclusion, harassment, and even physical violence. It can also undermine efforts to control the spread of the virus by creating barriers to effective communication, collaboration, and cooperation across different communities. Addressing hate speech in the COVID-19 pandemic requires a concerted effort by individuals, communities, and governments to promote accurate information, combat misinformation, and promote inclusivity and respect for diversity. This can include public education campaigns, regulations on hate speech, and community-based interventions that promote social cohesion and mutual understanding.

2) Discussion and Future Studies
As researcher mention earlier, this article found that out of 53 of the papers, only 20 of research and studies are being reviewed. This also conclude a word cloud and three main theme of those papers are revealed, which is 35% from the data is by theme Hate Speech Detection, 55% is literature from Hate Speech Issues and 10% are Hate Speech in Covid-19 Pandemic. The literature from 2018-2022 in Malaysia discusses the development and effectiveness of hate speech detection algorithms on social media platforms, such as Twitter and Facebook. Various approaches have been proposed, including machine learning techniques and rule-based methods, to identify hate speech content and reduce its spread.
Other than that, the literature also highlights the prevalence of hate speech on social media in Malaysia, which has been linked to rising levels of religious and racial tension. Scholars argue that hate speech online can fuel discrimination, prejudice, and even violence, and call for increased efforts to combat this problem.
Last but not least, theme of the data is hate speech in Covid pandemic. In the context of the Covid-19 pandemic, the literature discusses the emergence of hate speech related to the virus and its impact on marginalized communities in Malaysia. There have been reports of discriminatory and racist messages targeting specific groups, such as migrant workers and minorities, on social media platforms. Scholars emphasize the need for greater awareness and education to promote tolerance and understanding during this difficult time.
Overall, hate speech remains a significant problem in Malaysia, with various groups being targeted based on their ethnicity, religion, and political views. While the government has taken steps to combat hate speech, there is still a need for greater awareness, education, and enforcement to effectively address this issue.
For the future study, researcher suggest for other scholars to widen the setting of the topic, not only in Malaysia. As hate speech topic received worldwide attention, researchers can gather all the evidence of research and put in one paper on trends of hate speech. Scholar can put world map and point out which country has more studies on hate speech.
Other than that, the theme or topic of hate speech, researcher can add and diversify it into various topics such as ethnic and religious tensions, politically motivated hate speech, legal framework and enforcement and efforts to combat hate speech. Some of the topic not widely enough been research as hate speech actually a new issues in this information age.
Hence, there are a lot more to be study in this hate speech issues. Some of them never been touch by any of scholar in the world. This article reviewed the current literature on hate speech, reporting of hate speech trend research on hate speech on social media especially in Malaysia.

3) Contributions and Benefits of Study
Studying the trend of hate speech on social media in Malaysia can bring several benefits, including awareness and understanding of hate speech. Researching and analyzing the trend of hate speech on social media platforms can help create awareness about the prevalence, nature, and impact of hate speech in Malaysian society. It provides insights into the specific groups or topics that are targeted, the language used, and the underlying motivations behind hate speech. This knowledge enhances our understanding of the challenges posed by hate speech and facilitates informed discussions on potential solutions.
Other than that, this research can contribute a lot to policy development. A thorough study of hate speech trends can inform the development of effective policies and guidelines to address this issue. Policy-makers can utilize the findings to design legislation or regulations that encourage responsible online behavior, protect individuals and vulnerable communities from hate speech, and ensure the right to free speech is balanced with the need for a safe and inclusive digital environment.
Next, scholars also can benefit towards intervention and prevention of hate speech on social media. Understanding hate speech trends can aid in the development of intervention and prevention strategies. By identifying patterns and triggers, researchers can contribute to the design of targeted educational campaigns, awareness-raising initiatives, and community outreach programs. These efforts can promote digital literacy, tolerance, and respect, helping individuals recognize and counteract hate speech and its harmful effects.
Hate speech studies can contribute in building platform design and moderation. The findings of studying hate speech trends can guide social media platforms in enhancing their content moderation systems. By identifying common types of hate speech and the platforms or channels where it proliferates, researchers can offer recommendations on improving reporting mechanisms, automated detection algorithms, and community moderation practices. This can contribute to creating safer online spaces and reducing the dissemination of hate speech.
Ultimately, studying hate speech trends on social media in Malaysia serves the larger goal of fostering social cohesion and inclusion. By addressing hate speech and its underlying causes, researchers can contribute to building a more harmonious and tolerant society. Such efforts support the principles of diversity, equality, and respect for human rights, strengthening the fabric of Malaysian society. Overall, studying hate speech trends on social media in Malaysia brings valuable insights that can inform policies, interventions, and actions to combat hate speech, foster inclusivity, and promote a healthier online environment.