Data Profile for This Category
The typical data subjects for this category are government offices, government information disclosure offices, legal affairs divisions, and archives. Data sources include government information disclosure platforms, official archive systems, document management systems, departmental business ledgers, and relevant specifications of the Government Information Disclosure Regulations. Daily processed documents include government information disclosure application forms, supplementary notice letters, third-party opinion solicitation letters, reply letters, service receipts, and administrative reconsideration response materials. Data update frequency is irregular, with fields mainly covering applicant information, application matters, public attribute judgment basis, reply content, service records, etc. There is no fixed batch update rhythm.
Constraints Imposed on Knowledge Base Retrieval and Recall
Multi-source and scattered data sources require the retrieval link to adapt to government documents in different formats, and support semantic matching across system documents. The irregular update rhythm means that knowledge base synchronization must be triggered on demand to avoid redundant full updates. The case-oriented document structure requires retrieval to accurately match the specific application matters of applicants, and cannot rely on generalized keyword matching. The business requirements of ledger summarization and report generation require that recalled documents cover all relevant materials throughout the application process to ensure complete summarized information. At the same time, strict government affairs handling requirements mean that the accuracy of retrieval results directly affects the quality of subsequent replies, and irrelevant recalled content must be strictly controlled.
How to Set the Configuration
Reference only; verify against the target FastGPT release and deployment documentation.
| Configuration Item | Recommended Setting | Rationale |
|---|---|---|
similarity | 0.75–0.85 | Match precise application matter descriptions for government information application disclosure, filter irrelevant recalled content |
limit | Top 10 entries | Cover multiple related documents involved in a single case application, while keeping the context length within business requirements |
searchMode | embedding | Adapt to multi-source heterogeneous government documents, prioritize locating relevant content through semantic matching |
usingReRank | Enabled | Improve recall sorting accuracy for specific government documents such as reply letters and supplementary notice letters |
datasetSearchUsingExtensionQuery | Disabled | Avoid extended queries deviating from the specific application matters of applicants, ensure retrieval accuracy |
collectionIds | Configured according to department-specific knowledge bases | Achieve departmental data isolation, comply with government data security management requirements |
Three Common Configuration Errors
- Issue: An
invalid configuration parameter name "hnsw.iter"error is returned during retrieval. Cause: A configuration parameter name not supported by official FastGPT was mistakenly used, and the specified parameter list was not followed. - Issue: Unauthorized accounts can access documents in the department-specific knowledge base. Cause: The department-specific knowledge base collection was not configured via the
collectionIdsparameter, and data isolation was not implemented. - Issue: Insufficient retrieval results, failing to cover all related documents involved in the application. Cause: The
limitparameter value is too low, and multiple government documents required for a single case were not covered.
How to Confirm the Configuration Is Complete
- Enter the FastGPT knowledge base retrieval configuration interface, confirm that all used parameter names are within the officially supported range.
- Use a test account to initiate a simulated application retrieval, verify that recalled documents only come from the specific knowledge base collection bound to the current department.
- Simulate submission of multiple types of government applications, check that retrieval results cover document types such as application forms and reply letters.
- View retrieval logs to confirm that no error prompts for invalid configuration parameters appear.
The values provided are common starting points and should be measured against the reader's own samples.
Question material comes from public community discussions. Configuration values are common starting points and should be measured against your own samples. Verified on 2026-09-28.