Cazmireqo's Structured Review Capabilities
This page describes the tools and functions available through the Cazmireqo service interface, including classification, review, pattern detection, and output.
Core Classification and Review Tools
The following tools form the operational foundation of the Cazmireqo service. Each is described in terms of what it does and how users interact with it during a session.
Classification Engine
The classification engine processes incoming information and assigns category labels based on configurable rules. Rules can specify keyword patterns, structural signals, or relational characteristics between items in the same set.
The engine applies rules in sequence, allowing more specific conditions to override general ones when both match. Users with appropriate access can review the rule set applied during a session from within the interface configuration area.
Classification runs are deterministic for a given rule set. Running the same information through the same configuration will produce the same output. This predictability makes it easier to compare classification results across sessions.
Review Interface
The review interface presents classified information in a hierarchical panel layout. The top level of the panel shows major category groups. Selecting a group expands the panel to show items within that category, along with any sub-categories detected.
Users can move between categories without losing their position in the hierarchy. The interface preserves navigation state throughout a session, allowing users to return to previously viewed panels without reloading the classification output.
Each item panel shows the assigned label, the matching rule that triggered the classification, and any characteristics used in the decision. Items flagged as uncertain during classification are displayed with a distinct indicator.
From Submission to Reviewed Output
Submit Information
The user submits an information set to the service through the designated input channel. The format and size of the submission depend on the account configuration. The service acknowledges receipt and begins the classification process.
Classification Runs
The classification engine applies the configured rule set to the submitted information. Each item receives one or more category labels and is placed within the taxonomy. Items that do not meet any rule condition are grouped under an unclassified category for manual review.
Review and Export
The classified output becomes available in the review interface. Users navigate the category hierarchy, examine individual items, and retrieve session summaries. When the review is complete, users can request an organised output in the format supported by their configuration.
The Three Processing Stages
The diagram to the right illustrates the three stages every information set passes through when processed by Cazmireqo: input reception, classification, and review delivery. Each stage operates independently, which means the review interface remains usable while a subsequent classification run is in progress.
Users are notified when a new classification output is available, and can continue reviewing the previous session's output without interruption. This design supports users who process information sets on a recurring basis.
Additional Service Tools
Pattern Detection
Beyond assigning individual category labels, the service identifies recurring patterns across the information set. These may include items that consistently share characteristics across two or more categories, or groups of items that appear together in the input more frequently than would be expected.
Pattern detection results are surfaced in a separate section of the review interface. Users can examine detected patterns without navigating away from the main category hierarchy.
Structured Output
When a review session is complete, users can request a structured output of the classification results. This output includes the category hierarchy, item-level labels, pattern detection findings, and a session summary.
The format of the output depends on account configuration. Users can request output in text-based formats suitable for further use. The output does not include the original submitted information; it reflects only the classification results and detected structures.
Interested in Using the Service?
Submit an enquiry to the Cazmireqo support team to ask about access or to describe your intended use. The team reviews each submission and will follow up with relevant information.