Mobile App Reviews with Appraisal Theory Sentiment Labels

A manually labeled dataset of mobile app reviews for sentiment analysis. The data is annotated as High or Low performance based on appraisal theory parameters like insightfulness and practical usability. The dataset, authored by Ruping Zhang and last updated in June 2026, is stored in an XLS file of 9.5 KB.

Use Cases

• Training sentiment classification models based on the appraisal theory labels (High/Low).
• Benchmarking lightweight transformer models like MobileBERT for real-time analysis in resource-limited environments.
• Studying the application of contrastive learning (SimCLR) and data augmentation via WordNet for NLP feature extraction.
• Analyzing user feedback for specific app performance improvements based on parameters like transparency and consistency.

Strengths

• Labels are manually annotated and validated using the SHAP method of Explainable AI.
• The proposed model described achieved a mean fold accuracy of 88.63% and ROC AUC of 90.91%.
• Data is shared under a permissive CC-BY-4.0 license.

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• The dataset is very small at 9.5 KB, indicating limited scope.

Provenance

Source

figshare

Collection Method

Reviews were manually labeled based on performance parameters aligned with appraisal theory.

Time Range

null

Freshness

Last updated 2026-06-01 17:52:52; freshness should be verified.

Geography

null