Cruz, Christophe
Unlocking the Potential of Generative AI through Neuro-Symbolic Architectures: Benefits and Limitations
Bougzime, Oualid, Jabbar, Samir, Cruz, Christophe, Demoly, Frédéric
Neuro-symbolic artificial intelligence (NSAI) represents a transformative approach in artificial intelligence (AI) by combining deep learning's ability to handle large-scale and unstructured data with the structured reasoning of symbolic methods. By leveraging their complementary strengths, NSAI enhances generalization, reasoning, and scalability while addressing key challenges such as transparency and data efficiency. This paper systematically studies diverse NSAI architectures, highlighting their unique approaches to integrating neural and symbolic components. It examines the alignment of contemporary AI techniques such as retrieval-augmented generation, graph neural networks, reinforcement learning, and multi-agent systems with NSAI paradigms. This study then evaluates these architectures against comprehensive set of criteria, including generalization, reasoning capabilities, transferability, and interpretability, therefore providing a comparative analysis of their respective strengths and limitations. Notably, the Neuro > Symbolic < Neuro model consistently outperforms its counterparts across all evaluation metrics. This result aligns with state-of-the-art research that highlight the efficacy of such architectures in harnessing advanced technologies like multi-agent systems.
Enhancing Knowledge Graph Construction: Evaluating with Emphasis on Hallucination, Omission, and Graph Similarity Metrics
Ghanem, Hussam, Cruz, Christophe
Recent advancements in large language models have demonstrated significant potential in the automated construction of knowledge graphs from unstructured text. This paper builds upon our previous work [16], which evaluated various models using metrics like precision, recall, F1 score, triple matching, and graph matching, and introduces a refined approach to address the critical issues of hallucination and omission. We propose an enhanced evaluation framework incorporating BERTScore for graph similarity, setting a practical threshold of 95% for graph matching. Our experiments focus on the Mistral model, comparing its original and fine-tuned versions in zero-shot and few-shot settings. We further extend our experiments using examples from the KELM-sub training dataset, illustrating that the fine-tuned model significantly improves knowledge graph construction accuracy while reducing the exact hallucination and omission. However, our findings also reveal that the fine-tuned models perform worse in generalization tasks on the KELM-sub dataset. This study underscores the importance of comprehensive evaluation metrics in advancing the state-of-the-art in knowledge graph construction from textual data.
Political Events using RAG with LLMs
Arslan, Muhammad, Munawar, Saba, Cruz, Christophe
In the contemporary digital landscape, media content stands as the foundation for political news analysis, offering invaluable insights sourced from various channels like news articles, social media updates, speeches, and reports. Natural Language Processing (NLP) has revolutionized Political Information Extraction (IE), automating tasks such as Event Extraction (EE) from these diverse media outlets. While traditional NLP methods often necessitate specialized expertise to build rule-based systems or train machine learning models with domain-specific datasets, the emergence of Large Language Models (LLMs) driven by Generative Artificial Intelligence (GenAI) presents a promising alternative. These models offer accessibility, alleviating challenges associated with model construction from scratch and reducing the dependency on extensive datasets during the training phase, thus facilitating rapid implementation. However, challenges persist in handling domain-specific tasks, leading to the development of the Retrieval-Augmented Generation (RAG) framework. RAG enhances LLMs by integrating external data retrieval, enriching their contextual understanding, and expanding their knowledge base beyond pre-existing training data. To illustrate RAG's efficacy, we introduce the Political EE system, specifically tailored to extract political event information from news articles. Understanding these political insights is essential for remaining informed about the latest political advancements, whether on a national or global scale.
Sustainable Digitalization of Business with Multi-Agent RAG and LLM
Arslan, Muhammad, Munawar, Saba, Cruz, Christophe
Businesses heavily rely on data sourced from various channels like news articles, financial reports, and consumer reviews to drive their operations, enabling informed decision-making and identifying opportunities. However, traditional manual methods for data extraction are often time-consuming and resource-intensive, prompting the adoption of digital transformation initiatives to enhance efficiency. Yet, concerns persist regarding the sustainability of such initiatives and their alignment with the United Nations (UN)'s Sustainable Development Goals (SDGs). This research aims to explore the integration of Large Language Models (LLMs) with Retrieval-Augmented Generation (RAG) as a sustainable solution for Information Extraction (IE) and processing. The research methodology involves reviewing existing solutions for business decision-making, noting that many systems require training new machine learning models, which are resource-intensive and have significant environmental impacts. Instead, we propose a sustainable business solution using pre-existing LLMs that can work with diverse datasets. We link domain-specific datasets to tailor LLMs to company needs and employ a Multi-Agent architecture to divide tasks such as information retrieval, enrichment, and classification among specialized agents. This approach optimizes the extraction process and improves overall efficiency. Through the utilization of these technologies, businesses can optimize resource utilization, improve decision-making processes, and contribute to sustainable development goals, thereby fostering environmental responsibility within the corporate sector.
Knowledge Graph for NLG in the context of conversational agents
Ghanem, Hussam, Atmani, Massinissa, Cruz, Christophe
The use of knowledge graphs (KGs) enhances the accuracy and comprehensiveness of the responses provided by a conversational agent. While generating answers during conversations consists in generating text from these KGs, it is still regarded as a challenging task that has gained significant attention in recent years. In this document, we provide a review of different architectures used for knowledge graph-to-text generation including: Graph Neural Networks, the Graph Transformer, and linearization with seq2seq models. We discuss the advantages and limitations of each architecture and conclude that the choice of architecture will depend on the specific requirements of the task at hand. We also highlight the importance of considering constraints such as execution time and model validity, particularly in the context of conversational agents. Based on these constraints and the availability of labeled data for the domains of DAVI, we choose to use seq2seq Transformer-based models (PLMs) for the Knowledge Graph-to-Text Generation task. We aim to refine benchmark datasets of kg-to-text generation on PLMs and to explore the emotional and multilingual dimensions in our future work. Overall, this review provides insights into the different approaches for knowledge graph-to-text generation and outlines future directions for research in this area.
Imbalanced Multi-label Classification for Business-related Text with Moderately Large Label Spaces
Arslan, Muhammad, Cruz, Christophe
In this study, we compared the performance of four different methods for multi-label text classification using a specific imbalanced business dataset. The four methods we evaluated were fine-tuned BERT, Binary Relevance, Classifier Chains, and Label Powerset. The results show that fine-tuned BERT outperforms the other three methods by a significant margin, achieving high values of accuracy, F1-Score, Precision, and Recall. Binary Relevance also performs well on this dataset, while Classifier Chains and Label Powerset demonstrate relatively poor performance. These findings highlight the effectiveness of fine-tuned BERT for multi-label text classification tasks, and suggest that it may be a useful tool for businesses seeking to analyze complex and multifaceted texts.
Unlocking Insights into Business Trajectories with Transformer-based Spatio-temporal Data Analysis
Arslan, Muhammad, Cruz, Christophe
This approach allows us to not only analyze the performance of businesses over time but also understand how trends and performance vary across geographic territories. By combining data analysis with the latest advancements in natural language processing, we can gain a comprehensive view of business trends (Braşoveanu and Andonie 2020). It offers a powerful tool for unlocking insights into business trajectories, providing valuable information for businesses, investors, and policymakers. To perform business data analysis, we need to develop a news data analyzer (Alawadh et al. 2023). A news data analyzer refers to a system that processes and analyzes news articles to extract relevant information and insights (Lau et al. 2021).
Knowledge Base Approach for 3D Objects Detection in Point Clouds Using 3D Processing and Specialists Knowledge
Hmida, Helmi Ben, Cruz, Christophe, Boochs, Frank, Nicolle, Christophe
This paper presents a knowledge-based detection of objects approach using the OWL ontology language, the Semantic Web Rule Language, and 3D processing built-ins aiming at combining geometrical analysis of 3D point clouds and specialist's knowledge. Here, we share our experience regarding the creation of 3D semantic facility model out of unorganized 3D point clouds. Thus, a knowledge-based detection approach of objects using the OWL ontology language is presented. This knowledge is used to define SWRL detection rules. In addition, the combination of 3D processing built-ins and topological Built-Ins in SWRL rules allows a more flexible and intelligent detection, and the annotation of objects contained in 3D point clouds. The created WiDOP prototype takes a set of 3D point clouds as input, and produces as output a populated ontology corresponding to an indexed scene visualized within VRML language. The context of the study is the detection of railway objects materialized within the Deutsche Bahn scene such as signals, technical cupboards, electric poles, etc. Thus, the resulting enriched and populated ontology, that contains the annotations of objects in the point clouds, is used to feed a GIS system or an IFC file for architecture purposes.
Toward the Automatic Generation of a Semantic VRML Model from Unorganized 3D Point Clouds
Hmida, Helmi Ben, Cruz, Christophe, Nicolle, Christophe, Boochs, Frank
This paper presents our experience regarding the creation of 3D semantic facility model out of unorganized 3D point clouds. Thus, a knowledge-based detection approach of objects using the OWL ontology language is presented. This knowledge is used to define SWRL detection rules. In addition, the combination of 3D processing built-ins and topological Built-Ins in SWRL rules aims at combining geometrical analysis of 3D point clouds and specialist's knowledge. This combination allows more flexible and intelligent detection and the annotation of objects contained in 3D point clouds. The created WiDOP prototype takes a set of 3D point clouds as input, and produces an indexed scene of colored objects visualized within VRML language as output. The context of the study is the detection of railway objects materialized within the Deutsche Bahn scene such as signals, technical cupboards, electric poles, etc. Therefore, the resulting enriched and populated domain ontology, that contains the annotations of objects in the point clouds, is used to feed a GIS system.
From 9-IM Topological Operators to Qualitative Spatial Relations using 3D Selective Nef Complexes and Logic Rules for bodies
Hmida, Helmi Ben, Cruz, Christophe, Boochs, Frank, Nicolle, Christophe
This paper presents a method to compute automatically topological relations using SWRL rules. The calculation of these rules is based on the definition of a Selective Nef Complexes Nef Polyhedra structure generated from standard Polyhedron. The Selective Nef Complexes is a data model providing a set of binary Boolean operators such as Union, Difference, Intersection and Symmetric difference, and unary operators such as Interior, Closure and Boundary. In this work, these operators are used to compute topological relations between objects defined by the constraints of the 9 Intersection Model (9-IM) from Egenhofer. With the help of these constraints, we defined a procedure to compute the topological relations on Nef polyhedra. These topological relationships are Disjoint, Meets, Contains, Inside, Covers, CoveredBy, Equals and Overlaps, and defined in a top-level ontology with a specific semantic definition on relation such as Transitive, Symmetric, Asymmetric, Functional, Reflexive, and Irreflexive. The results of the computation of topological relationships are stored in an OWL-DL ontology allowing after what to infer on these new relationships between objects. In addition, logic rules based on the Semantic Web Rule Language allows the definition of logic programs that define which topological relationships have to be computed on which kind of objects with specific attributes. For instance, a "Building" that overlaps a "Railway" is a "RailStation".