What the data for this category looks like
Data for this category comes from the real estate registration information system, cadastral survey database, housing transaction online signing system, tax system, and electronic certificate repository. The update frequency is daily. Commonly processed documents include registration applications, ownership certificates, surveying and mapping results, tax payment vouchers, real estate ownership certificates and registers. Data fields cover right subject, real estate location, ownership source, restriction status, and more. There is no unified fixed unit; fields and units vary based on document type.
What constraints these characteristics impose on the "document parsing and chunking" link
Multi-source and heterogeneous document sources require adaptation to various formats of registration files, including scanned real estate ownership certificates, structured table-style surveying and mapping results, plain-text tax payment vouchers, and others. Core fields such as ownership source and restriction status must retain complete context; chunking must not split associated information. Daily updated batch data requires stable parsing and chunking processes to prevent single-file timeouts from affecting overall tasks. Errors in real estate registration are directly linked to administrative liability, so chunking must ensure key verification fields are not omitted or split incorrectly.
Configuration settings
These configurations apply to FastGPT v4.17.0:
| Configuration Item | Recommended Setting | Rationale |
|---|---|---|
chunkSize | 800–1200 characters | Adapts to the length of ownership descriptions and surveying and mapping data paragraphs in real estate registration documents, avoiding splitting core associated information |
chunkSplitter | Split by paragraph + title | Matches the structured title hierarchy of registration documents, preserving the contextual integrity of fields such as right subject and location |
customPdfParse | Enable enhanced parsing mode | Adapts to PDF documents with vector graphics or tables such as surveying and mapping results and real estate ownership certificates, extracting complete text and structured data |
UPLOAD_FILE_MAX_SIZE | 500 MB | Supports batch import of single-batch cadastral survey database export files, adapting to the file scale of daily business |
PARSE_FILE_TIMEOUT_SECONDS | 600 seconds | Addresses the parsing time required for large surveying and mapping result PDFs, preventing task failure due to timeout |
chunkSettingMode | Manual configuration | Adjust chunking rules for different types of registration documents, adapting to format differences between ownership certificates and tax payment vouchers |
Three common mistakes
- When uploading image-format registration documents such as scanned real estate ownership certificates, the parsing result is empty or only extracts a small amount of text. The cause is that the image OCR parsing capability is not enabled; the default parsing engine only supports plain-text format files.
- After deploying FastGPT locally and enabling the enhanced PDF parsing function, the locally deployed Docker-based MinerU service cannot be called. The cause is that the MinerU API access address and port are not filled in the FastGPT configuration, or the Docker port mapping is not correctly opened.
- The storage location of the knowledge base chunking and vectorization results cannot be found. The cause is that the mount directory of the FastGPT default vector database is not confirmed, or the custom storage path is not specified in the configuration file.
How to confirm the configuration is correct
- Upload a single surveying and mapping result PDF document, check whether the parsing result contains complete surveying and mapping coordinates and ownership description text, to verify whether the
customPdfParseconfiguration takes effect. - Upload a batch of registration application documents, check the completion rate of parsing tasks, to verify whether the
PARSE_FILE_TIMEOUT_SECONDSconfiguration adapts to the parsing time of large files. - View the vector database storage path on the FastGPT configuration page, confirm that it matches the actual mount directory, to verify whether the storage configuration is correct.
- Upload a scanned real estate ownership certificate image, check whether the parsing result contains the certificate number and right subject information, to verify whether the image OCR capability is enabled.
The values listed above are common starting points and should be evaluated 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.