The launch event of ARC Linkage Project, namely interaction mining for cyberbullying detection on social networks, is held in UTS in 1st Dec, 2017. This project is with four collaborators organizations, i.e., GBCA (Global Business College Australia), ARACY (Australia Research Alliance for Children & Youth), UQ (University of Queensland) and UTS (University of Technology Sydney). This collaboration is expected to increase technical capability for advanced data science application and build tools for social science researcher and psychologist in assisting them to effective detect cyberbullying activities and prevent the potential mental health damage of victims.

ARC linkage project

There are three student representations in this launch event as follow:

  • Detecting Abusive Texts on Social Networks;
  • Effective Conversation to Relieve Online User’s Mental Health Issue;
  • Privacy-aware Stochastic Optimization on Cyberbullying Detection.

My topic is about effective conversation to relieve online users’ mental health issue.

Following is the key point of my speech.

Cyberbullying and mental health is a global issue, especially severe in most developed countries and many emerging markets. Mental health issue is one of the most critical problems caused by cyberbullying. A person with mental health issue is usually bullied. And a person tends to develop mental health issue as a result of being bullied.

According to a survey, 52% young people report being cyber bullied. Bullying victims are 2 to 9 times more likely to consider committing suicide. According to a WHO report, 1 in 4 people worldwide suffer from mental disorder to some extent. And 3 out of 4 people with severe mental disorders do not receive treatment, which makes the problem worse. Partly due to severe mental disorder, 900,000 persons commit suicide each year all over the world, making suicide the second most common cause of death among the young.

A national survey showed that 35% of people with a mental disorder had used a health service and 29% consulted a GP within the 12 months before the survey.

Except visiting the psychologists, people suffering from mental health issue also use hotline, talk with social workers and get support from family friends and communities.

Among these types of treatment, talking is the key point of treatment to relieve mental health issue.

So far, online-conversation-based mental health services is likely to be the potential effective way to address the mental health issue. It provides an anonymous space and various forms of conversations. including patient to professional conversations, peer support conversations, patients to volunteers conversations and chatbot conversations. They could be one-to-one, one-to-many, and many-to-many conversations.

These forms of conversations are easy to access and people can get help from people all over the world. Also, online conversations is a more comfortable way comparing with face-to-face consultation.

Our aim is to use a automatic software called effective conversation assistant to help mental health patients to relieve their mental issue. Our software will be used by social workers and psychologists to improve their efficiency on producing effective online conversations.

There are three main functionalities of our effective conversation assistant. First, it can analyze the conversation data through natural language understanding techniques. Second, it can automatically understand the factors resulting in the mental health issue. Third, it evaluate the impact of sentences of the social workers might enter to the online conversations.


Feature Photo by unsplash-logoDan Meyers