Lightme: analysing language in internet support groups for mental health

  • PDF / 1,340,792 Bytes
  • 10 Pages / 595.276 x 790.866 pts Page_size
  • 48 Downloads / 192 Views

DOWNLOAD

REPORT


Health Information Science and Systems

RESEARCH

Lightme: analysing language in internet support groups for mental health Gabriela Ferraro1*  , Brendan Loo Gee2, Shenjia Ji3 and Luis Salvador‑Carulla4

Abstract  Background:  Assisting moderators to triage harmful posts in Internet Support Groups is relevant to ensure its safe use. Automated text classification methods analysing the language expressed in posts of online forums is a promising solution. Methods:  Natural Language Processing and Machine Learning technologies were used to build a triage post classi‑ fier using a dataset from Reachout.com mental health forum for young people. Results:  When comparing with the state-of-the-art, a solution mainly based on features from lexical resources, received the best classification performance for the crisis posts (52%), which is the most severe class. Six salient linguistic characteristics were found when analysing the crisis post; (1) posts expressing hopelessness, (2) short posts expressing concise negative emotional responses, (3) long posts expressing variations of emotions, (4) posts express‑ ing dissatisfaction with available health services, (5) posts utilising storytelling, and (6) posts expressing users seeking advice from peers during a crisis. Conclusion:  It is possible to build a competitive triage classifier using features derived only from the textual content of the post. Further research needs to be done in order to translate our quantitative and qualitative findings into fea‑ tures, as it may improve overall performance. Introduction Internet Support Groups (ISG) has been important and popular technologies for individuals with mental ill-health to receive support from other peers that have similar lived experiences[17] and to anonymously share their stories with others to support their recovery[26]. They are also referred to as online peer-support forum or networks. ISGs have supported groups of people with specific chronic health conditions, such as diabetes or mental health[17, 29]. Current evidence suggests ISGs may have a positive impact on individuals with mental illhealth; however, it may also exacerbate a person’s distress levels[19]. Nevertheless, the safe use of ISGs will require more attention, especially designing mechanisms that

*Correspondence: [email protected] 1 Commonwealth Scientific and Industrial Research Organization & Australian National University, GPO Box 1700, Canberra, ACT​2601, Australia Full list of author information is available at the end of the article © Springer Nature Switzerland AG 2020.

can assist in mitigating possible adverse effects and harm to ISG users[13]. Assessment and monitoring in ISGs are challenging and costly because it relies on the manual detection of posts in an online forum by trained moderators. This raises particular concerns on the scalability of ISGs as a potential digital health intervention. To overcome limitations, Natural Language Processing (NLP) and Machine Learning (ML) technologies can be used to build systems that can as