Summary of Mental Disorders Detection in the Era Of Large Language Models, by Gleb Kuzmin et al.
Mental Disorders Detection in the Era of Large Language Models
by Gleb Kuzmin, Petr Strepetov, Maksim Stankevich, Artem Shelmanov, Ivan Smirnov
First submitted to arxiv on: 9 Oct 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Artificial Intelligence (cs.AI)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This paper investigates the performance of traditional machine learning methods, encoder-based models, and large language models (LLMs) for detecting depression and anxiety. The study compares five datasets, each with varying formats and definitions of the target pathology class. AutoML models based on linguistic features, BERT, and state-of-the-art LLMs were tested as pathology classification models. The results show that LLMs outperform traditional methods, particularly on noisy and small datasets where training examples differ significantly in text length and genre. However, psycholinguistic features and encoder-based models can achieve performance comparable to language models when trained on texts from individuals with clinically confirmed depression. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper compares different ways computers can detect depression and anxiety using texts. It looks at five different sets of texts that were labeled as either having or not having depression or anxiety. The researchers tested three types of computer programs: traditional methods, programs based on language patterns (like BERT), and large language models. They found that the large language models are the best at detecting depression and anxiety, especially when the text is noisy or small. However, they also found that using specific words and phrases related to depression can be just as good as the language models. |
Keywords
» Artificial intelligence » Bert » Classification » Encoder » Machine learning
