What the data for this category looks like
Data sources for this category include the integrated budget management system, project declaration materials, fund payment data, business system ledgers, and satisfaction surveys. Data is updated annually. Commonly processed documents include performance target forms, self-assessment forms, performance evaluation reports, supporting material lists, and rectification reports. Data fields link performance targets to fund usage, and include performance target entries, fund disbursement records, supporting material numbers, rectification completion status, and other related content.
Constraints on citation sources and traceability
The annual update frequency requires the knowledge base to complete full data synchronization on a regular basis to avoid citing expired performance evaluation content. Multi-type document structures require adaptation to different chunking rules to ensure that key information such as performance targets, fund records, and rectification content is accurately split. Data from multiple system sources requires citation traceability to associate identifiers from different data sources, to avoid confusing supporting materials for different projects. Errors in this category are directly linked to budget adjustment and accountability mechanisms, requiring traceability results to precisely match the fields and content of original data, with no tolerance for vague associations or incorrect matches.
Configuration Settings
| Configuration Item | Recommended Setting | Rationale |
|---|---|---|
quoteMaxToken | 800–1200 token | Adapts to the single-paragraph length of performance evaluation documents, ensuring that cited content fully covers key information such as performance targets and fund records |
responseData | Enabled | Required to return associated identifiers of original data, used to link supporting materials from different systems |
detail | Enabled | Required to return the specific location of original documents, matching supporting material numbers in performance evaluation reports |
| Single recall citation count | Top 3 entries | Avoids overwhelming the reading experience of party building learning Q&A with excessive cited content |
| Citation timeout threshold | 60 seconds | Adapts to the average latency of multi-system data retrieval, preventing request interruptions |
Three Common Mistakes
- Phenomenon: When accessing the application via a no-login share link, citation sources are still displayed even if the citation display function is disabled. Cause: The front-end rendering switch for citation traceability was not disabled in the share configuration, causing the front-end to still load the citation module.
- Phenomenon: After upgrading to FastGPT v4.17.0, no citation identifier is displayed at the end of knowledge base answer paragraphs. Cause: The
responseDataconfiguration item was not enabled, or the citation display function was not enabled in the application's Q&A settings. - Phenomenon: After submitting multiple performance-related questions in succession, citation sources appear misaligned or incorrectly associated. Cause: Context-aware traceability configuration was not enabled, causing citations for new questions to not bind the performance project identifier for the current session.
How to Confirm Proper Configuration
- Submit a question targeting a single performance target, check if an associated citation identifier is displayed at the end of the answer paragraph.
- Click the citation identifier, verify that it jumps to the specific location of the corresponding original document and matches the supporting material number for the performance project.
- Generate a no-login share link, check that citation display can be toggled on or off via configuration switches after accessing the link.
- Submit two questions associated with different performance projects in succession, verify that citation sources are accurately bound to the document content for the corresponding project.
The values given 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.