What this category's data looks like
Typical entities for this category include central enterprise internal control departments, audit departments, finance departments, subordinate enterprises, and external audit institutions. Daily processed documents cover internal control self-evaluation reports, audit notices, audit working papers, issue lists, rectification ledgers, and rectification closure certificates. Data is mainly sourced from enterprise internal control systems, ERP, financial sharing systems, audit management systems, and procurement contract project systems. Update frequency is annual. Core time-consuming steps focus on extracting business samples and vouchers, forming audit issues and rectification requirements, tracking rectification evidence and closure processes.
What constraints these characteristics impose on the knowledge base retrieval and recall link
The annual updated data source feature means the knowledge base does not require frequent incremental synchronization. Only full data refresh once per year is needed to cover all business scenarios. Multi-source document types include structured fields and free text. Retrieval must balance precise matching of fixed fields and semantic understanding of free descriptions. The business nature of core time-consuming steps means retrieval must quickly locate historical issues and rectification cases. This avoids excessive operational time for audit personnel. At the same time, the consequences of errors in audit processes are severe. Retrieval results must have high precision, preventing low-relevance content from interfering with audit judgments.
How to set the configuration
Reference only; verify against the target FastGPT release and deployment documentation.
| Configuration Item | Recommended Value | Rationale |
|---|---|---|
similarity | 0.72–0.82 | Matches the precision requirement for audit issue descriptions, avoiding mixing in irrelevant rectification cases |
limit | Top 8 entries | Controls the number of returned results, adapting to the scenario where audit personnel quickly review historical cases |
searchMode | Semantic priority retrieval | Adapts to the mixed structure of this category's documents, which include both structured fields and free text |
usingReRank | Enabled | Optimizes sorting results, preventing key rectification cases from being overshadowed by low-relevance content |
datasetSearchUsingExtensionQuery | Enabled | Expands the retrieval scope with synonymous expressions, covering common issue synonyms used in audits |
collectionIds | Specify the collection IDs bound to audit management and internal control systems | Limits retrieval data sources, excluding redundant non-category business data |
Three common mistakes
- Issue: Insufficient retrieval precision, returning a large number of irrelevant rectification ledger entries. Cause: No reasonable
similaritythreshold set, orusingReRankfunction not enabled, leading to low-relevance results being included in the retrieval list. - Issue: Local images in imported Markdown documents fail to display properly. Cause: Local file paths are directly used as image links, without uploading images to the unified storage path supported by the platform.
- Issue: Exported knowledge base files have small size and incomplete content. Cause: Only metadata for partial collections is exported, full associated data source collections are not selected, or complete chunked content is not checked during export.
How to confirm the configuration is correct
- A typical audit issue description may be submitted, with the relevance of returned results reviewed and the
similaritythreshold adjusted to a range that meets business requirements. - A Markdown test document containing local images may be uploaded, with correct loading and display of images verified.
- The full knowledge base file may be exported, with the match between the file size and the total capacity of the data sources checked, and confirmation obtained that all target collections are covered.
- The retrieval API may be called, with the number of returned results verified to match the value set for
limit.
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.