SleepDepNet Evaluation Metrics for Sleep and Depression Classification from Reddit

Akshi Kumar's 2026 dataset contains evaluation metrics for the SleepDepNet model, a multi-task learning framework analyzing user-generated text. The dataset, stored in an XLS file of 5.5 KB, includes performance scores for classifying sleep quality and depressive sentiment from Reddit posts. Experimental results reported include an F1-score of 0.89 for sleep quality and 0.86 for depressive sentiment analysis.

Use Cases

• Benchmarking transformer-based models for sleep quality classification based on the reported F1-score of 0.89
• Evaluating multi-task learning frameworks for joint sentiment and sleep analysis based on the described SleepDepScore
• Analyzing linguistic patterns associated with mental health in online discourse based on the described attention mechanisms and emotion-aware features

Strengths

• Includes specific, high-performance evaluation metrics (F1-scores of 0.89 and 0.86) for two classification tasks
• Released under a permissive CC-BY-4.0 license for reuse and redistribution
• Model framework incorporates interpretability features like attention mechanisms, as described

Limitations

• Row count is unknown, which may limit suitability assessment
• Column-level documentation is absent; field semantics must be inferred after download
• The 5.5 KB file size suggests a very limited scope, likely containing only summary metrics rather than raw data

Provenance

Source

figshare

Collection Method

Derived from experimental results of the SleepDepNet model applied to text data from Reddit communities.

Time Range

null

Freshness

Last updated 2026-05 17:37:13

Geography

null