SafeGuard: A Web-Based Application to Guard Against Cyberbullying

In this day and age there is a consistent rise in the usage of social media amongst all age groups thus, more people have established an online presence that makes these users more susceptible to becoming a target of cyberbullying. SafeGuard is a web-based application that monitors social media accounts and detects critical keywords such as “hurt”, “die”, etc., and phrases to detect cyberbullying threats and emotional distress in a person’s social media post. Once a monitor discovers a match, an alert is generated for the end user in which they can initiate human intervention. This application is a new strategy for academic institution that allows them to be proactive rather than reactive against cyberbullying.


Introduction
Traditional physical bullying happens in playgrounds, work area and other aspects of daily life. In such cases the victim come in physical contact either verbally or physically with their bully. There is a widespread public concern regarding bullying among youth because of substantial impact on healthy and development. as indicated in published research (Eriksen, 2018;Eslea & Smith, 2012;Fekkes, Pijpers, & Verloove-Vanhorick, 2005;Papanikolaou, Chatzikosma, & Kleio, 2011;Tenenbaum, Varjas, Meyers, & Parris, 2011). Most of the interaction among younger generation these days is online where they express their emotions and ideas, such exchanges can be emotionally stressful for some thus causing negative impact on their personalities and mental development. The advancement of information technology started seeing a new form of bullying called cyber bullying. It is also referred to as electronic bullying, cyberbullying, or online social cruelty (Kowalski & Limber, 2007). The National Crime Prevention Council of Canada (Paullet & Pinchot, 2014) defines cyberbullying as "the process of using the Internet, cell phones or other devices to send or post text or images intended to hurt or embarrass another person." One also needs to understand that both cyberbullying and physical bullying have many things in common however, cyberbullying has distinct difference. Cyberbullying occurs when youth repeatedly use technology to threaten or harass their peers (Paullet & Pinchot, 2014).
In the 21st century as a society we spend a lot of time on one of the social media platforms and wake up to the tweets on Twitter, instead of newspaper. Among these users there are millions of youths, that post sensitive and disturbing content and are susceptible to cyberbullying, etc. Authors (Schneider, O'Donnell, Stueve, & Coulte, 2012;Lenhart, Maddeen, & Hitlin, 2005) report 93% of youth are active users of the Internet and 75% own a cell phone thus there is great potential for cyberbullying among these users. In cyberbullying, the offenders can hide his/her identity (Donegan, 2012) and can remotely contact the victim in numerous ways anonymously using any of the electronic social medium, thus may feel diminished responsibility and accountability as compared with traditional bullying (Juvonen & Gross, 2008;Mishna, Saini, & Solomon, 2009). In addition, the victim can be harassed round the clock by using modern technology which can otherwise be used to improve our quality of life. Kowalski and Limber (2007) conducted a study of 168 undergraduate students to understand student's perception of cyberbullying including their knowledge, opinion, and personal experiences if they have any. The paper reports 66% have witnessed and 50% of the students believe that cyberbullying has become a normal part of life. The paper also reports what participants of the study think needs to be done to avoid cyberbullying in future. In addition, authors in (Stacey, 2009) investigated the issues of cyberbullying and report on how students are coping with this issue through discussion with 74 students. Since students in academic institution have access to cyberworld via internet including computers and mobile phones. The research in the paper states that social networking sites and synchronous chat sites are places where cyberbullying most commonly occur, with email and texting on mobile phones.
The frequency of suicide among youth victims of cyberbullying at national level (Schneider, O'Donnell, Stueve, & Coulte, 2012;Lenhart & Hitlin, 2005) has raised concerns thus many states in USA now have legislation in place that requires academic institutions to have policies to address cyberbullying in their antibullying polices. However, academic institution may lack the ability or knowhow on how to detect cyberbullying and identify which youths are offenders and which are victims before it causes any physical or mental damage to the individual. Bullying in general is a problem among youth sometimes as a victim or as a bully or even both as well as not limited to academic institutions (Burk, DiRenzo, & Fuller, 2019). It is because of rising violence in academic institutions in USA and cyberbullying amongst youth we have conclude that there is an immediate need to develop an application to address this situation.

Proposed Application
Recently, there has been a rise in mass shootings, cyberbullying, and suicide cases in academic institutions, where a good portion of youth usually post on their social media platform before committing to any of these acts. If someone in the school management can identify such individuals as well their post in a timely manner, then there is a possibility that prospective victim can be protected and saved from any physical and/or mental harm. The goal is to tackle the imminent danger of school shootings, suicides, and cyberbullying by carefully monitoring student's social media activities for any mentions of suicide, death, violence, etc. and sending an alert to the school officials in a timely manner. Thus, we are proposing an application that will monitor social media postings, utilizes web technologies, and is hosted in a cloud environment that is accessible from any location. It provides seamless integration with an academic institution's existing security solutions that will allow for better threat and emotional distress detection of the students using Twitter in the academic institution. The intent of this application is to alert users and institution's administration before a tragedy occurs by looking at suspicious signs on students' online posts. Using this web-based application, provides users a way to create and maintain lists of student accounts as well as monitor their social media text exchanged or sent that will be used to detect potential threats or emotional distress by parsing posts from a student's social media account, like Twitter. By monitoring social media posts, SafeGuard will be able to determine if the language used, matches any preset keywords or phrases. This will generate an alert in the system if a condition is met and will notify the end user of the system. It will require an initial setup and minimal maintenance to achieve a high level of monitoring, the content will be filtered down to only alert about specific posts which would otherwise not be possible if done by human monitoring alone. This greatly decreases the amount of time and personnel effort needed to monitor online media posts among student population. Though the primary focus of this product is that detection of a violent act or identifying a person in emotional distress can prevent tragic incidents occurring such as suicide, shooting or cyberbullying.

Application Requirement
A major benefit is to catch an issue before it becomes something big and/or tragic. The institution and law enforcement will anticipate the incident through our alerts and will be more prepared and might even be able to prevent that tragic incident from occurring. It will not only benefit other innocent people involved in the tragedy, but also the youth who might be suffering from pain and anger from being bullied by getting them the help they need with the help of more informed officials. The academic institute management need to be educated on ensuring that they are maintaining up to date student lists, monitors, and users to ensure their data size does not grow larger than it needs to be which could result in slower queries. Another concern with data size and querying if the creation of monitors is very generic then in turn it will generate too many notifications that would make the client need to sort through more notification than needed.
In addition, the client must be aware of physical security of the devices, maintaining users in the system, creating monitors, reading notifications, maintaining student information in the institutions and ensure they are protected from unauthorized use. They will be responsible for creating policies and directives around their use of the application and will not be aided by SafeGuard other than providing access to the technology for them to use. The institution needs to whitelist their desired IP addresses with the application so access to the application is limited to those specific devices. Passwords must be strong and meet basic requirements such as lowercase, uppercase, numbers, special characters, and a minimum length. An access control restricted by IP address is used for the client's access by their network, JSON web token is used that has an expiration. All clients using the application requires authentication and authorization on each request. The application uses the HTTPS protocol for client requests.       ication that is ts for human i lowing an inst lity by adding words more th p". By adding fore. earning page e used to monit intervention. M titution to be p more features han others suc g weights, it w tor social med Monitors can b roactive rather s such as addin ch as the keyw would help the Vol. 10,No. 4; dia posts and d be set up to d r than reactive ng weightage f word "rape" ha e system filte 2021 detect detect . The factor aving r out