iASN: Indonesian Civil Service Chatbot Query Routing Dataset

336 labeled text samples for training a chatbot routing model for the iASN civil service system in Sumenep, Indonesia. The dataset was created by author moh-affan and was last updated on June 4, 2026. It is formatted as CSV with text and label columns.

Use Cases

• Training a query router for a civil service chatbot based on the three defined intents (RAG, AGENT, CHITCHAT).
• Benchmarking text classification models on a small, domain-specific Indonesian language dataset.
• Developing intent detection systems for employee information portals based on the described query types.
• Studying patterns in public service inquiries based on the labeled categories of regulation, personal data, and general conversation.

Strengths

• Contains 336 labeled samples, providing a defined scale for model training.
• Labels are clearly defined into three distinct categories (RAG, AGENT, CHITCHAT) relevant to the application domain.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• The dataset's small size (336 samples) may limit the complexity of models it can effectively train.

Provenance

Source

huggingface

Collection Method

Created for training a chatbot routing model for the iASN (BKPSDM Sumenep) system.

Freshness

Last updated 2026-06-04 02:15:50; freshness should be verified.

Geography

Sumenep, Indonesia (implied by the mention of BKPSDM Sumenep).