Fine-Tuned LLM Utilities for Psychological Questionnaire Generation

169 professional psychological questionnaires were used to fine-tune large language models for assessment tasks. The optimized Qwen-2.5 and GLM-4 models showed gains in text generation quality, logical consistency, and cultural adaptability. Zhitao Yuan published this 66.5 KB dataset under a CC-BY-4.0 license in April 2026.

Use Cases

• Fine-tuning language models for psychological text generation based on the corpus of 169 questionnaires.
• Evaluating model performance on scientific rigor and logical consistency metrics as described in the study.
• Developing culturally adaptive psychological assessment tools based on the cross-regional expression conversion feature.
• Benchmarking text generation quality improvements (BLEU-4, ROUGE-L) for specialized NLP tasks.

Strengths

• Dataset is derived from a corpus of 169 professional psychological questionnaires.
• Model optimization demonstrated specific gains: BLEU-4 increased by 0.05 and ROUGE-L by 0.057.
• Logical consistency of generated text was improved by 28.6%.
• Cross-regional expression conversion achieved over 85% accuracy.

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 66.5 KB, indicating limited scope.

Provenance

Source

figshare

Collection Method

Specialized fine-tuning scheme applied to a corpus of psychological questionnaires, integrating instruction fine-tuning with human feedback reinforcement.

Time Range

Publication date is 2026, but the temporal coverage of the underlying questionnaire corpus is unknown.

Freshness

Last updated 2026-04-24 17:29:22; freshness should be verified.

Geography

The study mentions cross-regional expression conversion, but the specific geographic scope of the data is not stated.