Understanding AgileIM’s Data Rules Engine and AI Rule Behaviors: Classification vs. Filing

In AgileIM, the Data Rules Engine offers versatile tools to manage documents efficiently, leveraging AI rules for classification and filing. While both types of rules—classification and folder filing—may seem similar due to shared terminologies, their underlying mechanisms and objectives differ significantly, leading to different outcomes even for seemingly identical inputs. This paper explores these distinctions and highlights opportunities to streamline processes.

1. AI Rule for Classification

The classification AI rule is designed to determine the appropriate category or classification for a document based on its content. This process uses:

• Classification Definitions: These include classification names, codes, and scope notes (e.g., “3900 BYLAWS-01: General” or “0110 ADMINISTRATION – GENERAL-01”).
• Scope Notes: Contextual guidelines that define what types of documents belong to a classification.

For example, the rule might identify a document discussing local government bylaws as belonging to 3900 BYLAWS-01 based on its content and scope alignment.

2. AI Rule for Filing

The folder filing AI rule determines where a document should be placed within a predefined folder structure, such as a file plan like the BC LGMA structure. This rule focuses on:

• Folder Paths: Hierarchically organized folders (e.g., `/0100-699 ADMINISTRATION/0110 [ADMINISTRATION – GENERAL]/0110-01 [General]`).
• Filing Context: Document properties and potential folder matches based on names and metadata.

Here, the document may be filed under a folder matching its classification, but the engine might also propose new folders if the existing structure doesn’t adequately cover the document’s content.

3. Why Results Differ

Despite overlapping elements like names or codes, these two AI rules yield different results due to distinct objectives:

• Classification Rules: Aim to categorize documents for governance, retention, or compliance purposes based on universal or contextual definitions.
• Filing Rules: Focus on practical organization within a specific operational structure, which may not align perfectly with classification frameworks.

For instance, a document classified as 3900 BYLAWS-01 might be proposed for a folder `/0100-699 ADMINISTRATION/0110 [ADMINISTRATION – GENERAL]/0110-01 [General]` because the folder structure prioritizes administrative context over specific bylaws.

4. The Role of Inheritance

An efficient solution to harmonize these processes lies in classification inheritance:

• Applying the classification to a folder allows all documents in that folder to inherit the classification automatically.
• This approach eliminates the need to run both classification and filing rules independently for the same document, saving time and reducing inconsistencies.

For example, if `/0100-699 ADMINISTRATION/0110 [ADMINISTRATION – GENERAL]/0110-01 [General]` is associated with 3900 BYLAWS-01, any document filed there would automatically be classified under that category.

Conclusion

The apparent divergence between classification and filing outcomes stems from their distinct roles in document management. By leveraging classification inheritance, organizations can streamline processes, ensuring that documents are both properly categorized and efficiently organized. This approach reduces redundancy, enhances accuracy, and simplifies user workflows within AgileIM.