Artificial intelligence (AI) can detect loneliness with 94 percent accuracy based on a person's language, reports a new scientific paper.
Researchers in the United States used various AI tools, including IBM Watson, to analyze transcripts from older adults who were asked about feelings of loneliness.
By analyzing words, phrases, and gaps in silence during the interviews, the AI rated the symptoms of loneliness almost as accurately as the loneliness questionnaires completed by the participants themselves, which can be biased.
It found that lonely people tend to have longer answers to direct questions about loneliness and express more sadness in their answers.
A team led by researchers from the University of California's San Diego School of Medicine used artificial intelligence technologies to analyze natural language patterns (NLP) and detect levels of loneliness in older adults
"Most studies use either a direct question," How often do you feel lonely? "What can lead to biased responses from the stigma associated with loneliness," said senior writer Ellen Lee of the UC San Diego (UCSD) School of Medicine.
For this project, we used natural language processing, an unbiased quantitative assessment of the emotions and feelings being expressed, along with the usual tools to measure loneliness.
"What's interesting about the tool is that it not only uses a dictionary-based approach – for example, finding specific words that reflect fear – but also as a template for the words used in the answer."
There has been a "loneliness pandemic" in the United States, characterized by rising suicide and opioid use rates, lost productivity, increased health care costs, and rising mortality rates in the United States.
WHAT IS NATURAL VOICE PROCESSING?
Natural Language Processing (NLP) is a branch of AI that helps computers understand, interpret, and manipulate human language.
NLP helps computers communicate with people in their own language and scales other language-related tasks.
With NLP, for example, computers can read text, hear speech, interpret it, measure mood, and determine which parts are important.
A UC San Diego study published earlier this year found that 85 percent of residents in an independent senior living community reported moderate to severe loneliness.
The Covid-19 pandemic and the resulting lockdowns have increased the amount of time people have been in solitude and made the situation worse.
Researchers wanted to know more about how natural language processing techniques and machine learning models can predict loneliness of older adults in shared apartments.
The study focused on 80 independent seniors between the ages of 66 and 94 years with a mean age of 83 years.
Before the pandemic, trained study staff conducted semi-structured interviews with participants between April 2018 and August 2019.
Participants were asked 20 questions from the UCLA Loneliness Scale, which is a four-point rating scale for questions such as "How often do you feel left out?" Uses. and & # 39; how often do you feel like you are part of a group of friends? & # 39;
None of the questions in the UCLA Loneliness Scale explicitly use the word "lonely".
Participants were also interviewed in face-to-face conversations that were recorded and manually transcribed.
The transcripts were then examined using natural language processing tools, including IBM's WNLU (Watson Natural Language Understanding) software, to quantify the mood and the emotions expressed.
WNLU uses deep learning to extract metadata from keywords, categories, sentiments, emotions and syntax.
Participants completed semi-structured interviews on experiencing loneliness and a self-reporting scale (UCLA loneliness scale) to rate loneliness, which were then compared. The transcripts were fed into the IBM Watson Natural Language Understanding program (pictured).
"Natural language patterns and machine learning allow us to systematically examine long interviews of many people and examine how subtle language features such as emotions can indicate loneliness," said lead author Varsha Badal of UCSD.
"Similar analysis of human emotions would be biased, inconsistent, and require extensive training to standardize."
Using linguistic characteristics, the AI was able to predict loneliness compared to the “quantitative model” – the results of the UCLA loneliness scale – with an accuracy of 94 percent.
Lonely individuals had longer answers in face-to-face interviews and expressed greater sadness when answering direct questions about loneliness.
The study also found differences between men and women – the latter were more likely than men to report feeling lonely during the interviews.
And men used more fearful and joyful words than women in their responses, suggesting that their experiences with negative and positive emotions were more extreme or that men were more free to express those emotions.
"There can be subtle gender differences in mood and emotion in how older men and women describe feeling lonely in response to a direct question," Lee told MailOnline.
The study reveals discrepancies between research ratings for loneliness and an individual's subjective experience of loneliness that could be identified by AI.
There might be "lonely language" that could be used to identify loneliness in older adults, say the researchers.
With IBM Watson, users can parse text to extract metadata from content such as concepts, entities, keywords, categories, feelings, emotions, relationships, and semantic roles using natural language understanding
This could improve doctors and families' assessments and management of loneliness in older adults, especially during social isolation.
UCSD is now examining natural pattern signatures of loneliness and wisdom that are inversely linked in older adults, i.e. when one rises, the other falls.
"Voice data can be combined with our other assessments of cognition, mobility, sleep, physical activity, and mental health to improve our understanding of aging and promote successful aging," said Dilip Jeste, co-author of the study at UCSD.
The study measured the accuracy of the AI using participants' own reports of loneliness, which, as noted, do not always reflect true feelings and emotions.
However, AI and self-reports can be used together by psychologists and professionals to increase the accuracy of a diagnosis.
"We agree that the score (UCLA Loneliness Scale) has some inaccuracies because it is based on the self-report," Lee told MailOnline.
'However, it is one of the most popular tools in research because it does not use the word' lonely 'explicitly and seems to consistently capture the trait of loneliness without prejudice by gender.
"We hope to develop more sensitive tools to assess the state of loneliness."
The study was published in the American Journal of Geriatric Psychiatry.
HOW TO LEARN ARTIFICIAL INTELLIGENCES WITH NEURAL NETWORKS
AI systems are based on Artificial Neural Networks (ANNs) that try to simulate how the brain works in order to learn.
ANNs can be trained to recognize patterns in information – including speech, textual data, or visual images – and are the foundation of a large number of developments in AI in recent years.
Traditional AI uses inputs to "teach" an algorithm on a particular subject by feeding it vast amounts of information.
AI systems are based on Artificial Neural Networks (ANNs) that try to simulate how the brain works in order to learn. ANNs can be trained to recognize patterns in information – including speech, textual data, or visual images
Practical uses include Google's voice translation services, Facebook's face recognition software, and Snapchat's live image manipulating filters.
Entering this data can be very time consuming and is limited to one type of knowledge.
A new generation of ANNs called Adversarial Neural Networks pit the minds of two AI bots against each other, allowing them to learn from each other.
This approach aims to accelerate the learning process and refine the results generated by AI systems.
(tagsToTranslate) dailymail (t) sciencetech