Distribution of Slot Tagset in Annotated Persian Dialogue Data

Pegah Safari published a dataset on figshare in May 2026 detailing the distribution of slot tagsets in annotated data. The data supports research on profile extraction from Persian dialogue systems, a less-resourced language. The dataset is 5.5 KB in size and is available under a CC-BY-4.0 license.

Use Cases

• Benchmarking slot-filling models for Persian dialogue systems based on the described annotated data.
• Training hybrid profile extraction systems combining slot filling, in-context learning, and ontology-based inference as described.
• Evaluating model performance on detecting inconsistencies in user-provided information, a task mentioned in the description.
• Comparing the performance of large language models against specialized methods for low-resource languages.

Strengths

• The associated research paper reports a high F-score of 90.46 for the proposed method.
• The method's inconsistency detection accuracy of 92% substantially exceeds reported performances of GPT-4o and Llama-3-70B.
• The dataset is explicitly focused on Persian, addressing a gap in resources for less-resourced languages.

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• The description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

figshare

Collection Method

Likely annotated from dialogue systems as part of a research project on profile extraction.

Freshness

Last updated 2026-05-11 17:27:38; freshness should be verified.

Geography

Focus is on Persian language data.