Research Articles — PROTEX
RESEARCH ARTICLES — STRUCTURED KNOWLEDGE AND CONTROLLED AI SYSTEMS

This page collects selected research papers developed within the PROTEX project.

The publications focus on how knowledge can be structured, retrieved, and used within AI systems operating under explicit constraints and defined boundaries.

The work reflects an ongoing investigation into controlled AI architectures, structured retrieval systems, and the relationship between knowledge representation, decision support, and system behaviour.

Selected publications
Quantitative Evaluation of Native Microsoft Copilot Studio on the PROTEX Behavioural Homicide Corpus

A 200-question benchmark evaluating factual retrieval, comparative reasoning, false-premise handling, uncertainty preservation, and semantic contamination resistance within a structured behavioural knowledge corpus.

Migration to Native Microsoft Copilot Studio

A study examining retrieval stability, knowledge structure, uncertainty preservation, and behavioural consistency following migration from a custom retrieval architecture to native Microsoft Copilot Studio.

RAG System Design

A methodological exploration of retrieval-augmented generation systems based on structured knowledge access and controlled context.

Hybrid Retrieval Architecture

A study of combined semantic and deterministic retrieval approaches, focusing on control, filtering, and contextual precision.

Case Description Framework

A framework for structuring complex case material into distinct analytical layers, preserving the separation between fact, interpretation, and narrative.

Procedural AI Assistants

A study of AI systems designed for procedural environments, focusing on bounded roles, structured knowledge, and decision support.

Context

These publications form the methodological foundation for system directions explored across the PROTEX project, including procedural AI systems, expert knowledge interfaces, structured retrieval architectures, and Microsoft Copilot knowledge environments.

They are intended as reference points for understanding how AI systems can operate not as open-ended generators, but as controlled systems grounded in structured knowledge, explicit constraints, evidentiary boundaries, and measurable reliability.

Przewijanie do góry