PGMOnto2021: First International Workshop on Joint Use of Probabilistic Graphical Models and Ontology
Context and Objective:
An ontology is well known to be the best way to represent knowledge in a domain of discourse. It is defined by Gruber as “an explicit specification of a conceptualization”. It allows to represent explicitly and formally existing entities, their relationships, and their constraints in an application domain. This representation is the most suitable and beneficial way to resolve many challenging problems related to information domain (e.g., semantic interoperability among systems, knowledge sharing, and knowledge capitalization). Ontology formalization can be based on First order logic (FOL) to describe concepts, relationships, and constraints, enabling it to make inferences and giving it a graphical representation. Using ontology has many advantages, among them we can cite ontology reusing, reasoning and explanation, commitment, and agreement on a domain of discourse, ontology evolution and mapping, etc.
Over the last three decades, graphical probabilistic models (PGMs) have enjoyed a surge of interest as a practically feasible framework of expert knowledge encoding and as a new comprehensive data analysis framework.
Probabilistic graphical models (PGMs) such as Bayesian network, influence diagram or probabilistic relational model are considered as one of the most successful tools that enable a compact representation of complex system and the increased ability to effectively learn and perform inference in large networks. Besides the compact representation of probability, PGMs are also intuitively easier for a human to understand than joint probabilities because they highlight the direct dependences between random variables and their overall semantics is easily captured visually through their graphical representation.
In practice, the combination of PGMs and ontologies might be beneficial to have high expressiveness and reasoning possibilities under uncertainty. Despite the difference between these two domain representation models, they have the potential to complement each other: part of the value of ontology baseline knowledge may be used to enhance PGM by resolving challenging tasks: (i) the identification of relevant variables (variable selection), (ii) the determination of structural relationships between the considered variables (arcs), and (iii) the estimation of parameters associated to the model. Once the PGM is learned, its results can be used together with ontology reasoning engine to perform probabilistic inference.
This first regular workshop aims at demonstrating recent and future advances in Semantic Probabilistic Graphical Models and Probabilistic Ontologies. Moreover, this workshop offers an invaluable opportunity to boost collaboration and conversation between Industrial Experts and academic researchers, allowing therefore ideas exchanging and presenting results of on-going research in structured knowledge and causality approaches.
This regular workshop is well closed to the relevant topics of the conference KGSWC-2021: https://kgswc.org/ such as Knowledge Graphs, Ontology and Reasoning, Causal Reasoning, etc.
We have organized an international workshop about Deep Learning meets Ontology and Natural Language Processing (https://sites.google.com/view/deepontonlp-eswc2021), a national workshop related to deep learning techniques for NLP tasks: (https://www.dl-nlpegc2021.ml/accueil) and special sessions in an international Conference on Machine Learning and Applications (ICMLA-2020).
We invite submission of papers describing innovative research and applications around the following topics. Papers that introduce new theoretical concepts or methods, help to develop a better understanding of new emerging concepts through extensive experiments, or demonstrate a novel application of these methods to a domain are encouraged.
Topics of interests:
- Construction of probabilistic ontologies
- Construction of semantic probabilistic graphical model
- Semantic causality and probability
- Causality and ontology
- PGM for ontology mapping
- PGM learning
- Ontology for PGM construction
- Probabilistic inference engine
- Tools, systems and applications
The workshop is open to submit unpublished work resulted from research that present original scientific result, methodological aspects, concepts and approaches. All submissions must be PDF documents written in English and formatted according to KGSWC-2021 format https://www.springer.com/gp/computer-science/lncs/conference-proceedings-guidelines . Page limits are inclusive of references and appendices, if any. Papers are to be submitted through the Easychair Conference Management System. Please note that paper submissions are anonymous.
We welcome the following types of contributions:
- Full research papers (8-10 pages): Finished or consolidated R&D works, to be included in one of the Workshop themes.
- Short papers (5-6 pages): Ongoing works with relevant preliminary results, opened to discussion.
At least one author of each accepted paper must register for the workshop, in order to present the paper, there, and to the conference. For further instructions please refer to the KGSWC-2021 page (https://kgswc.org/).
- Workshop paper submission due: October 01, 2021
- Workshop paper notifications: October 23, 2021
- Workshop paper camera-ready versions due: November 02, 2021
- Workshop: November 19-20, 2021 (half-day)
- Estimation of the number of attendees: 50
Sarra BEN ABBES, Research Scientist, ENGIE-France: firstname.lastname@example.org
Ahmed MABROUK, Research Scientist, ENGIE-France: email@example.com
Lynda TEMAL, Research Scientist, ENGIE-France: firstname.lastname@example.org
Philippe Calvez, Research Lab Manager, ENGIE-France: email@example.com
Stefan FENZ, key researcher at SBA Research and Senior Scientist at Vienna University of Technology Philippe LERAY, Professor at University of Nantes
Estimation of the audience size: >=40
International Workshop on Joint Use of Probabilistic Graphical Models and Ontology