Artificial Intelligence To Help Predict Levels of Loneliness

(Photo: pixabay)

According to experts, there has been an increase in the number of people who experience loneliness, evidenced by the rising suicidal rates and opioid use. The COVID-19 pandemic and implementation of social distancing and lockdowns have only made things worse.

Researchers at the University of California San Diego School of Medicine used artificial intelligence technologies to analyze speech or natural language patterns (NLP) among older adults to predict their level of loneliness. This information was published online in the American Journal of Geriatric Psychiatry.

Natural language is an umbrella term surrounding various techniques that process or analyze unstructured natural speech and text. With advancements in artificial intelligence and machine learning systems, several conditions may be detected just by analyzing a person's natural speech. 

READ MORE: Kansas Researchers Develop Self-Monitoring App for Student Productivity

Ellen Lee, MD, senior author of the study and an assistant professor of psychiatry at UC San Diego School of Medicine, said that most studies asked the question "how often do you feel lonely?" which can lead to biased responses. In this study, researchers made use of natural language processing so that they can achieve an unbiased quantitative assessment of expressed emotion and sentiment.

Older adults are the most vulnerable group to suffer from loneliness. According to the UC San Diego study, 85 percent of adults living in an independent senior housing community have reported moderate to severe levels of loneliness.

In this new study, 80 participants aged 66 to 94, with a mean age of 83 years. But aside from asking questions and documenting answers to the UCLA Loneliness Scale, participants were also evaluated in more unstructured interviews lasting up to 90 minutes. This was then analyzed using NLP-understanding software.  

Varsha Badal, Ph.D., a postdoctoral research fellow and first author of the study, said that using NLP can allow researchers to examine long interviews and explore how subtle speech features like emotions may indicate loneliness. If humans analyzed these emotions, it would likely be biased or inconsistent.

The AI system could qualitatively predict a subject's loneliness with up to 94 percent accuracy. The system found out that the lonelier a person feels, the longer their responses were to direct questions regarding loneliness. The researchers even suggest that the "lonely speech" pattern could be used to monitor the well-being of older subjects. 

The Difference In How Men and Women Express Loneliness

We all know that men and women handle negative emotions differently. Men can express joyful or fearful words or express anger when they don't feel good about themselves, while women explicitly vocalize feelings of loneliness.

A study conducted by Shelley Borys at the University of Waterloo found that women may not necessarily feel lonelier but are more comfortable admitting they're lonely because of less negative consequences. In contrast, men are more reluctant due to their effects on their masculinity.

The researchers were optimistic that complex AI systems could intervene in real-time and help individuals reduce their loneliness by adopting positive cognitions, managing social anxiety, and engaging in meaningful social activities.

ALSO READ ABOUT: Joint Study Finds Parental Touch Reduces Babies' Response to Pain 

Check out more news and information on Psychology on MD News Daily.