What This Category of Data Looks Like
Data for this category originates from the Chinese Center for Disease Control and Prevention Information System, medical institution HIS and infectious disease reporting systems, and laboratory testing systems. The update frequency is daily. Daily documents include infectious disease report cards, epidemic analysis reports, case correction records, clustered outbreak reports, and epidemiological investigation reports. Most document structures include mandatory fields required for public health statutory reports, such as basic case information, diagnosis results, reporting time limits, and correction instructions. Relevant content must be processed in strict compliance with public health reporting specifications.
What Constraints These Characteristics Impose on Document Parsing and Chunking
Data sources are a mix of structured documents and unstructured reports exported from business systems. Batch documents updated daily require consistent parsing. The completeness requirement for statutory report fields means key information must be accurately captured during parsing, to avoid splitting field associations during chunking. Different document types, such as report cards and analysis reports, have large format differences, so different chunking logic must be adapted. Contextual association information required for case correction and duplicate card verification must not be lost after chunking, to ensure accurate and usable information for subsequent summary and early warning workflows.
Configuration Settings
Reference only; verify against the target FastGPT release and deployment documentation.
| Configuration Item | Recommended Value Range | Rationale |
|---|---|---|
chunkSize | 800–1200 characters | Adapts to the field length of infectious disease report cards and epidemic analysis reports, avoiding splitting of key statutory report fields |
chunkSplitter | Split by statutory report fields | Matches the fixed structure of report fields within documents, retains contextual association required for case correction and duplicate card verification |
customPdfParse | Enable structured parsing mode | Adapts to formatted PDF documents exported from disease control systems, accurately extracting tables and fixed field content |
UPLOAD_FILE_MAX_SIZE | 1000 MB | Supports batch uploading of multiple daily summary public health report files |
PARSE_FILE_TIMEOUT_SECONDS | 600 seconds | Adapts to the parsing time required for long-text epidemic analysis reports, avoiding mid-task interruptions for batch jobs |
chunkSettingMode | Manual configuration | Adjusts chunking rules for different document types, meeting the retention requirements for statutory report fields |
Three Common Errors
- After uploading a PDF report exported from a disease control system, the data processing result is empty, and no parsing completion prompt is displayed in the interface. Cause: The
customPdfParsestructured parsing mode is not enabled, so formatted PDF table content exported from the system cannot be recognized. - After batch uploading multiple infectious disease report cards, some documents lose correction record fields after parsing and chunking. Cause:
chunkSplitteris configured to split by regular paragraphs, and does not match the statutory report field structure, resulting in split field associations. - After uploading a long-text epidemic analysis report, the parsing task times out and fails. Cause: The value of
PARSE_FILE_TIMEOUT_SECONDSis too low, and does not adapt to the parsing time required for long-text documents.
How to Verify Successful Configuration
- Upload a single standardized infectious disease report card PDF, view parsed chunked content, and confirm that key fields such as basic case information and diagnosis results are not split.
- Batch upload 3 to 5 daily public health report documents, check the completion status of the data processing queue, and confirm that there are no batch parsing failure records.
- Trigger a search test to verify that correction records and clustered outbreak-related content in the report card can be recalled.
- Upload a summary epidemic analysis report larger than 500 MB, and confirm that no timeout error occurs during the parsing process.
The values provided are common starting points and should be evaluated using samples from actual deployment contexts.
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.