Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Neuro-Conceptual Artificial Intelligence: Integrating OPM with Deep Learning to Enhance Question Answering Quality

Xin Kang, Veronika Shteingardt, Yuhan Wang, Dov Dori

TL;DR

The paper addresses the challenge of explainable QA by marrying neural learning with symbolic reasoning through Neuro-Conceptual Artificial Intelligence (NCAI). It proposes a framework that uses Object-Process Methodology (OPM) as a conceptual backbone, mapping natural language to OPM models via in-context learning and performing QA with an OPM-grounded reasoning process (OPM-QA). Key contributions include the NCAI framework, the OPM-QA system, and quantitative transparency metrics that measure alignment between predicted reasoning and OPM logic. Experimental results on a 50-question, multi-hop dataset show that OPM-QA yields higher accuracy and substantially better transparency than a NL-based QA baseline, demonstrating the value of explicit conceptual modeling for explainable neuro-symbolic QA. The work offers a scalable path toward more interpretable AI systems and provides open-source data and code for future benchmarking and extension.

Abstract

Knowledge representation and reasoning are critical challenges in Artificial Intelligence (AI), particularly in integrating neural and symbolic approaches to achieve explainable and transparent AI systems. Traditional knowledge representation methods often fall short of capturing complex processes and state changes. We introduce Neuro-Conceptual Artificial Intelligence (NCAI), a specialization of the neuro-symbolic AI approach that integrates conceptual modeling using Object-Process Methodology (OPM) ISO 19450:2024 with deep learning to enhance question-answering (QA) quality. By converting natural language text into OPM models using in-context learning, NCAI leverages the expressive power of OPM to represent complex OPM elements-processes, objects, and states-beyond what traditional triplet-based knowledge graphs can easily capture. This rich structured knowledge representation improves reasoning transparency and answer accuracy in an OPM-QA system. We further propose transparency evaluation metrics to quantitatively measure how faithfully the predicted reasoning aligns with OPM-based conceptual logic. Our experiments demonstrate that NCAI outperforms traditional methods, highlighting its potential for advancing neuro-symbolic AI by providing rich knowledge representations, measurable transparency, and improved reasoning.

Neuro-Conceptual Artificial Intelligence: Integrating OPM with Deep Learning to Enhance Question Answering Quality

TL;DR

The paper addresses the challenge of explainable QA by marrying neural learning with symbolic reasoning through Neuro-Conceptual Artificial Intelligence (NCAI). It proposes a framework that uses Object-Process Methodology (OPM) as a conceptual backbone, mapping natural language to OPM models via in-context learning and performing QA with an OPM-grounded reasoning process (OPM-QA). Key contributions include the NCAI framework, the OPM-QA system, and quantitative transparency metrics that measure alignment between predicted reasoning and OPM logic. Experimental results on a 50-question, multi-hop dataset show that OPM-QA yields higher accuracy and substantially better transparency than a NL-based QA baseline, demonstrating the value of explicit conceptual modeling for explainable neuro-symbolic QA. The work offers a scalable path toward more interpretable AI systems and provides open-source data and code for future benchmarking and extension.

Abstract

Knowledge representation and reasoning are critical challenges in Artificial Intelligence (AI), particularly in integrating neural and symbolic approaches to achieve explainable and transparent AI systems. Traditional knowledge representation methods often fall short of capturing complex processes and state changes. We introduce Neuro-Conceptual Artificial Intelligence (NCAI), a specialization of the neuro-symbolic AI approach that integrates conceptual modeling using Object-Process Methodology (OPM) ISO 19450:2024 with deep learning to enhance question-answering (QA) quality. By converting natural language text into OPM models using in-context learning, NCAI leverages the expressive power of OPM to represent complex OPM elements-processes, objects, and states-beyond what traditional triplet-based knowledge graphs can easily capture. This rich structured knowledge representation improves reasoning transparency and answer accuracy in an OPM-QA system. We further propose transparency evaluation metrics to quantitatively measure how faithfully the predicted reasoning aligns with OPM-based conceptual logic. Our experiments demonstrate that NCAI outperforms traditional methods, highlighting its potential for advancing neuro-symbolic AI by providing rich knowledge representations, measurable transparency, and improved reasoning.

Paper Structure

This paper contains 23 sections, 5 equations, 3 figures, 5 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Neuro-Symbolic AI Approaches
  4. Interpretability and Transparency in Language Models
  5. Language Models and Knowledge Graphs for Question Answering
  6. NCAI Framework
  7. Object-Process Methodology for NCAI
  8. Converting Natural Language to OPM using In-Context Learning
  9. OPM Knowledge-Based Question-Answering System
  10. Experiments
  11. Experiment Setup
  12. Data:
  13. Knowledge Sources:
  14. QA Systems:
  15. Evaluation Metrics:
  16. ...and 8 more sections

Figures (3)

  • Figure 1: Overview of the NCAI framework, illustrating how the LLM converts natural language text into structured OPM knowledge and uses it in OPM-QA for transparent reasoning. Starting from the text "Heuristic often starts as an informal rule of thumb …", the model generates an OPM model and answers questions by referencing processes like Heuristic-to-Principle Evolving.
  • Figure 2: Constructed OPDs illustrating the transformation of a Heuristic from a rule of thumb to a principle through various OPM elements—processes, objects, and states—within the OPM framework.
  • Figure 3: In-Zoomed Diagram (SD1) highlighting the specific processes in blue involved in transforming Heuristic from documented & shared to theoretically backed. These highlighted processes match exactly those identified by OPM-QA in Table \ref{['tab:answer_comparison']}, demonstrating a coherent and transparent reasoning path.