Conversational Analysis Agents for Depression Detection: A Systematic Review

Abstract

Depression is known as a non-cognitive disturbance that can be seen among different people all over the world. This pertains to disorders that have affected cognitions and behaviors that arise from overt disorders in cerebral function. It is more common for young adults to elderly people based on lifestyles, work pressure, personal problems, diseases, people who had strokes or hemorrhages, certain brain diseases, and paralysis. This paper is focused on reviewing the research papers previously done on detecting depression. Utilizing predefined search systems, we have gone through a couple of studies zeroing in on gloom and involved conversational information for location and conclusion. The objective of this research is to review large research studies on whether conversational agents can detect and diagnose depression by using smart texting analysis. The study was done by searching IEEE Xplore, Sci-hub, Doi, Scopus, and Pubmed using a predefined search strategy. This review was focused on studies that include the possibilities and steps of detecting depression and diagnosis that involved conversational data or analysis agents after assessing them by independent reviewers and relevancy for eligibility. After retrieving more than 117 references initially it was narrowed down to 95 references that were found relevant as most of them applied analytical techniques and technology-based solutions. Detecting depression and diagnosing it through smart texting analysis is a broad and emerging field and has a promising future but not every research studies were robust enough to get valid results in the end. This study aimed to keep the review as precise and informative as possible.