MATIH Platform is in active MVP development. Documentation reflects current implementation status.
14. Context Graph & Ontology
Analytics & Ranking
Analytics Overview

Analytics Overview

The Context Graph analytics subsystem provides intelligent entity suggestions, pattern mining across agent traces, decision precedent ranking, and feedback-driven quality scoring. These services work together to surface insights from the knowledge graph and improve agent performance over time.


Subsections

PageDescription
Entity SuggestionsAuto-completion and typeahead suggestions for entities, tags, and queries
Pattern MiningSequential and graph-based pattern discovery from agent traces
Decision RankingMulti-factor ranking of past decisions for precedent search
Feedback ScoringMulti-factor trace quality assessment from explicit and implicit feedback

How Analytics Fit Together

User Query --> Entity Suggestions (typeahead)
                    |
            Search Results --> Decision Ranking (precedent scoring)
                    |
            Agent Execution --> Pattern Mining (pattern discovery)
                    |
            User Feedback --> Feedback Scoring (quality signals)
                    |
            Improved Rankings and Suggestions (feedback loop)

Key Services

ServiceSourcePurpose
EntitySuggestionServiceservices/entity_suggestion_service.pyFast prefix-based and semantic suggestions
PatternMiningServiceservices/pattern_mining_service.pySequential pattern and anomaly detection
DecisionRankingServiceservices/decision_ranking_service.pyMulti-factor precedent ranking
FeedbackScoringServiceservices/feedback_scoring_service.pyComposite trace quality scoring