Logic Augmented Generation

Semantic Knowledge Graphs (SKG) face challenges with scalability, flexibility, contextual understanding, and handling unstructured or ambiguous information. However, they offer formal and structured knowledge enabling highly interpretable and reliable results by means of reasoning and querying. Larg...

Full description

Saved in:
Bibliographic Details
Main Authors: Aldo Gangemi, Andrea Giovanni Nuzzolese
Format: Article
Language:English
Published: Elsevier 2025-05-01
Series:Web Semantics
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S1570826824000453
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850054055217332224
author Aldo Gangemi
Andrea Giovanni Nuzzolese
author_facet Aldo Gangemi
Andrea Giovanni Nuzzolese
author_sort Aldo Gangemi
collection DOAJ
description Semantic Knowledge Graphs (SKG) face challenges with scalability, flexibility, contextual understanding, and handling unstructured or ambiguous information. However, they offer formal and structured knowledge enabling highly interpretable and reliable results by means of reasoning and querying. Large Language Models (LLMs) may overcome those limitations, making them suitable in open-ended tasks and unstructured environments. Nevertheless, LLMs are hardly interpretable and often unreliable. To take the best out of LLMs and SKGs, we envision Logic Augmented Generation (LAG) to combine the benefits of the two worlds. LAG uses LLMs as Reactive Continuous Knowledge Graphs that can generate potentially infinite relations and tacit knowledge on-demand. LAG uses SKGs to inject a discrete heuristic dimension with clear logical and factual boundaries. We exemplify LAG in two tasks of collective intelligence, i.e., medical diagnostics and climate projections. Understanding the properties and limitations of LAG, which are still mostly unknown, is of utmost importance for enabling a variety of tasks involving tacit knowledge in order to provide interpretable and effective results.
format Article
id doaj-art-5ca1d0bf93a243f9b8d900358ec49886
institution DOAJ
issn 1570-8268
language English
publishDate 2025-05-01
publisher Elsevier
record_format Article
series Web Semantics
spelling doaj-art-5ca1d0bf93a243f9b8d900358ec498862025-08-20T02:52:23ZengElsevierWeb Semantics1570-82682025-05-018510085910.1016/j.websem.2024.100859Logic Augmented GenerationAldo Gangemi0Andrea Giovanni Nuzzolese1University of Bologna, Bologna, Italy; CNR - Institute of Cognitive Sciences and Technologies, Bologna, ItalyCNR - Institute of Cognitive Sciences and Technologies, Bologna, Italy; Corresponding author.Semantic Knowledge Graphs (SKG) face challenges with scalability, flexibility, contextual understanding, and handling unstructured or ambiguous information. However, they offer formal and structured knowledge enabling highly interpretable and reliable results by means of reasoning and querying. Large Language Models (LLMs) may overcome those limitations, making them suitable in open-ended tasks and unstructured environments. Nevertheless, LLMs are hardly interpretable and often unreliable. To take the best out of LLMs and SKGs, we envision Logic Augmented Generation (LAG) to combine the benefits of the two worlds. LAG uses LLMs as Reactive Continuous Knowledge Graphs that can generate potentially infinite relations and tacit knowledge on-demand. LAG uses SKGs to inject a discrete heuristic dimension with clear logical and factual boundaries. We exemplify LAG in two tasks of collective intelligence, i.e., medical diagnostics and climate projections. Understanding the properties and limitations of LAG, which are still mostly unknown, is of utmost importance for enabling a variety of tasks involving tacit knowledge in order to provide interpretable and effective results.http://www.sciencedirect.com/science/article/pii/S1570826824000453Knowledge graphsLarge language modelsLogic augmented generation
spellingShingle Aldo Gangemi
Andrea Giovanni Nuzzolese
Logic Augmented Generation
Web Semantics
Knowledge graphs
Large language models
Logic augmented generation
title Logic Augmented Generation
title_full Logic Augmented Generation
title_fullStr Logic Augmented Generation
title_full_unstemmed Logic Augmented Generation
title_short Logic Augmented Generation
title_sort logic augmented generation
topic Knowledge graphs
Large language models
Logic augmented generation
url http://www.sciencedirect.com/science/article/pii/S1570826824000453
work_keys_str_mv AT aldogangemi logicaugmentedgeneration
AT andreagiovanninuzzolese logicaugmentedgeneration