Skip to main content

2024 | Buch

Artificial Intelligence for Knowledge Management, Energy and Sustainability

10th IFIP International Workshop on Artificial Intelligence for Knowledge Management, AI4KMES 2023, Krakow, Poland, September 30–October 1, 2023, Revised Selected Papers

herausgegeben von: Eunika Mercier-Laurent, Gülgün Kayakutlu, Mieczyslaw Lech Owoc, Abdul Wahid, Karl Mason

Verlag: Springer Nature Switzerland

Buchreihe : IFIP Advances in Information and Communication Technology

insite
SUCHEN

Über dieses Buch

This volume IFIP AICT 693 constitutes the refereed proceedings of the 10th IFIP International Workshop on Artificial Intelligence for Knowledge Management, AI4KMES 2023, from September 30th – October 1st, 2023, held in Krakow, Poland.

The 15 full papers presented together with 2 short papers were carefully reviewed and selected from 49 submissions. The accepted papers covered a large scope of topics related to sustainability in various contexts such as smart cities, agriculture, energy and gas production and distribution, industry, management and biodiversity.

Inhaltsverzeichnis

Frontmatter
Artificial Intelligence, Sustainability and Climate Change
Abstract
This paper presents a short introduction to the AI4S workshop aiming in painting a “big picture” of various interrelated aspects of Sustainability and Climate Change. The connections between energy, agriculture, water and technology, usually addressed separately, are vital to properly address the given challenges.
The concern about the sustainability and climate change weakly considers in the debates the pollution induced by technology and innovation. While technology produces twofold effect – benefits and waste, the capacity of available technology and in particular Artificial Intelligence for related complex problems solving is still underexplored.
The knowledge-based AI and connectionist approaches and techniques combined with adequate thinking may innovate the way of managing industry, administration and in other contexts and improve the effectiveness of efforts in sustaining the Planet. Knowledge-based AI help improving the explanation of the results got from Deep learning.
After the introduction aiming in connecting the “dots” between AI, sustainability and Climate actions, this paper presents shortly some contribution of AI to Climate actions and comments the conditions for minimizing the footprint.
Eunika Mercier-Laurent
In Search for Model-Driven eXplainable Artificial Intelligence
Abstract
This paper reports on ongoing and innovative research in the area of eXplainable Artificial Intelligence (XAI). A classical XAI task is considered as finding an explanation of the model generated via Machine Learning by identifying the most influential variables for local decision-making. Such an approach suffers from severe limitations. The proposed approach moves the explanatory process to a new, knowledge-level dimension. It is oriented towards Model Discovery, i.e. the internal structure and functions of the components. The concept of Model-Driven XAI is put forward and explained with examples. An experiment on Function Discovery via Grammatical Evolution is reported in brief.
Antoni Ligęza, Dominik Sepioło
An Intelligent Chatbot Based on Hybrid Approach Implementing Technical Knowledge to Support Maintenance and Training Activities at Electricity of France (EDF)
Abstract
From a practical point of view, Knowledge Management (KM) is both a mix of well-established corporate procedures and an emerging field of academic research. The opposition might be strong between these different points of view. To go a step forward, we had a pragmatic approach (Lorino 2016; Thievenaz 2019), including theory and practice, engineers, and academics. We will see how an innovation can germinate at low noise level, with a classical KM problem: experts are retiring soon, how could we “transfer” their experience?
Therefore, we followed a Data Sliming methodology (Kayakutlu & Mercier-Laurent 2016), combining Technical Expertise and Contextual Data. In this paper, we focus on an air conditioner: how and where to install it? The discussions in the team (made up of engineers, technicians, and schematic designers, then computer scientists and, to a lesser extent, academics) set up an innovation: a “conversationalist”, i.e., a chatbot whose dialog is that of an expert.
In the first part, we will present the air conditioner problem and the problem of semantic shift. In power plants, “long term Knowledge” (Dourgnon; Mercier-Laurent; Roche 2005) has been alive for decades! In the second part, we will see how the team built “design sheets” for the air conditioner. We will follow the discussions and see how we made a living chatbot.
This innovation was made possible by a favourable environment, with the help of the Open Innovation Team, and has been hailed as such.
Anne Dourgnon, Alain Antoine, Pascal Albaladejo, Fabrice Vinet
Supporting Food Computing with Ontologies and Artificial Intelligence Methods for Sustainability
Abstract
This paper investigates the potential of combining food ontologies and AI in the food sector for enhanced sustainability. We argue that this combination can foster sustainable food systems, underscoring how semantic structures and AI can facilitate precision agriculture, sustainable food choices, personalized diets, and climate change mitigation. Our goal is to discuss how these innovative technologies can be harnessed to better understand, manage, and ultimately transform the food domain for a sustainable future. As a first step towards achieving this goal, we provide an overview of prominent food ontologies and knowledge graphs in the food domain highlighting their structures and focal points, and we illustrate the value of ontological reasoning through practical food domain examples, using SPARQL queries and ontological reasoning for insightful knowledge derivation. We also discuss how to combine AI and ontologies to create new knowledge resources for improved data integration and management.
Weronika T. Adrian, Julia Ignacyk, Katarzyna Pyrczak, Krzysztof Kluza, Piotr Wiśniewski, Antoni Ligęza
The Use of Semantic Networks for the Categorization of Prosumers: Expanded Version
Abstract
Business continuity is possible through maintaining market position, while growth requires gaining competitive advantage. This can only be achieved through systematic attention to the consumer and the development of the product offering. These factors primarily determine the success of an organization. Against this backdrop, the trend towards personalization of goods delivered to consumers is becoming increasingly evident. The answer to these business problems is the creation of buyerseller relationships in which the consumer becomes a prosumer: a consumer who provides opinions, suggests solutions, tests and evaluates the advantages and disadvantages of the product. Therefore, the search for methods and tools that allow for the identification of groups of consumers susceptible to cooperation with the organization - to varying degrees - becomes a very important scientific and business problem. This is why the authors of this article defined the aim of the article as the analysis of the possibility of using semantic networks for the categorization of consumers and defining the category of prosumers. The results presented in the article were obtained through the application of triangulation of research methods, such as analytical literature review, computer simulation conducted using Protégé software, and research experiment consisting of simulation of selected problem situations. This paper is an expanded version of an article published as part of the ECAI 2023 conference proceedings.
Iwona Chomiak-Orsa, Andrzej Greńczuk, Kamila Łuczak, Dorota Jelonek
Methods for Mitigating Gender Bias in Binary Classification Models – A Comparative Analysis
Abstract
Inequality is one of the problems of the modern world. Discrimination of various kinds can affect many areas of life. The growing importance of data in the modern world makes it all the more important to ensure that the methods used to analyze it do not return results in which unfairness is present. Unfortunately, there may be situations where there is unfairness in the predictions of machine learning models. Over the years, researchers in this field of artificial intelligence have developed methods for mitigating bias in models. The purpose of this article is to identify gender bias in selected dataset and compare which of the selected solutions achieves the best result while used for mitigating impact of this type of unfairness on model’s predictions. The following research methods were used: literature review, experiment and comparative analysis. The evaluation of methods will be based on the value of measures: disparity in recall and disparity in selection rate for the column containing information about the person’s gender. The values of these measures, achieved by binary classification models in which different methods for mitigating bias were implemented, will be compared in order to identify which of the methods is best suited for mitigating gender bias in binary classification models.
Andrzej Małowiecki, Iwona Chomiak-Orsa
ChatGPT as a Learning Tool in Business Education. Research on Students’ Motivation
Abstract
The education process is radically changing with changing technology, which fits perfectly into sustainable development. However, how the teachers and students use it depends only on them. The effectiveness of ChatGPT as a learning tool in business education and its impact on student motivation can vary depending on implementation, context, or individual student characteristics. Using ChatGPT as a learning tool in business education can be an innovative approach to engage students and enhance their learning experience. In our research, we would like to create a diagnosis regarding the use of ChatGPT by WUE students (motives, scope of questions/tasks, frequency of use, concerns, satisfaction). The impact of AI (ChatGPT) on the implementation of knowledge verification processes of students at economic universities and the resulting demands for changes in the learning outcomes verification system - diagnosis of an economic university (UEW). Without a doubt, further research and empirical studies are needed to explore these aspects and provide more comprehensive insights.
Aleksander Binsztok, Beata Butryn, Katarzyna Hołowińska, Mieczysław L. Owoc, Małgorzata Sobińska
Integrating Artificial Intelligence into Electric Vehicle Energy Systems: A Survey
Abstract
Electric vehicles play an important role in the global transition to “Net Zero” and the decarbonisation of point source emissions from road transport, as their market share continues to grow each year. However, integrating these electric vehicles into existing electricity grids, refuelling stations and supply chains presents significant challenges. But most challenges can be effectively addressed through the implementation of artificial intelligence solutions. This paper starts with an introduction to artificial intelligence and an investigation of the challenges associated with the widespread adoption of electric vehicles in society. Then, the paper analyses in detail how artificial intelligence can facilitate the smooth integration of electric vehicles. In addition, the paper provides a detailed investigation of the life-cycle emissions of electric vehicles and examines the limitations of existing research in this area and suggests potential avenues for future research.
Weiqi Hua, Daniel Mullen, Abdul Wahid, Khadija Sitabkhan, Karl Mason
Towards Sustainable Power Systems: Exploring the Opportunities of Multi-task Learning for Battery Degradation Forecasting
Abstract
The task of developing a reliable and long-lasting battery system remains a major obstacle to the expansion of the electric vehicle fleet and the improvement of storage systems in renewable energy plants. To overcome this challenge, improving battery ageing models to accurately forecast the battery degradation trajectory is a complex but critical step. In this work, we introduce a novel sequence-to-sequence multi-task learning (MTL) method. Empirical results on battery ageing datasets show that our model enhances the data efficiency and precision of capacity and power degradation forecasts for lithium-ion battery cells. Furthermore, our model generates new insights into the trade-offs and key decisions for incorporating battery ageing models into the MTL paradigm.
Emilie Grégoire, Sam Verboven
Classification Tree Based AI System for Short Term Prediction for Heat and Power Plants
Abstract
Companies supplying electrical energy rely mainly on long term agreements with electricity producers, but on the other hand the actual demand should be precisely predicted for 48 h ahead, to take into account the actual weather conditions. The same type of analysis is important also for heat and power plants, but this time the temperature of returning water is the most interested. Some time series models can be used for forecasting. However in daily practice – more popular are average profiles showing the distribution over 24 h. We propose to build an AI system to choose the future profile. First – from the historical data – daily profiles are obtained, by cutting the time series into 24-h periods. Then, these empirical profiles are clustered with hierarchic and non-hierarchic clustering procedures to form homogeneous groups (types of profiles). Finally the classification methods are applied using weather data and observed demand from previous days (up to one week backwards). The measure for the forecasting evaluation has been proposed. Out of the two tested classification methods, CART classification tree performed better.
Małgorzata Markowska, Andrzej Sokołowski, Grzegorz Migut, Danuta Strahl
The Influence of Neural Networks on Hydropower Plant Management in Agriculture: Addressing Challenges and Exploring Untapped Opportunities
Abstract
Hydropower plants are crucial for stable renewable energy and serve as vital water sources for sustainable agriculture. However, it is essential to assess the current water management practices associated with hydropower plant management software. A key concern is the potential conflict between electricity generation and agricultural water needs. Prioritising water for electricity generation can reduce irrigation availability in agriculture during crucial periods like droughts, impacting crop yields and regional food security. Coordination between electricity and agricultural water allocation is necessary to ensure optimal and environmentally sound practices. Neural networks have become valuable tools for hydropower plant management, but their black-box nature raises concerns about transparency in decision making. Additionally, current approaches often do not take advantage of their potential to create a system that effectively balances water allocation.
This work is a call for attention and highlights the potential risks of deploying neural network-based hydropower plant management software without proper scrutiny and control. To address these concerns, we propose the adoption of the Agriculture Conscious Hydropower Plant Management framework, aiming to maximise electricity production while prioritising stable irrigation for agriculture. We also advocate reevaluating government-imposed minimum water guidelines for irrigation to ensure flexibility and effective water allocation. Additionally, we suggest a set of regulatory measures to promote model transparency and robustness, certifying software that makes conscious and intelligent water allocation decisions, ultimately safeguarding agriculture from undue strain during droughts.
C. Coelho, M. Fernanda P. Costa, L. L. Ferrás
Usability of Honeybee Algorithms in Practice. Towards Nature-Inspired Sustainable Development
Abstract
Honeybee algorithms (HBAs), inspired by the collective intelligence of bee colonies, have garnered increasing attention for their potential to solve complex problems (especially focused on optimization aspects). This abstract delves into the practical usability of these algorithms, exploring their strengths and limitations in real-world applications, particularly within the context of sustainable development. Advantages of the approaches in problem solutions relate to efficiency (HBAs demonstrate impressive performance in finding optimal solutions, particularly for large-scale and dynamic problems), adaptability (their decentralized nature allows them to handle changes in the environment effectively, and sustainability (often require fewer resources compared to traditional optimization methods, contributing to a more sustainable approach). Some examples of implementation in sustainable development embrace resource optimization: which can optimize energy usage in buildings, transportation networks, and industrial processes, waste management: which can aid in designing efficient waste collection and recycling systems, and renewable energy integration: optimizing the placement and operation of renewable energy sources like solar panels and wind turbines.
The article, aimed at popularizing bee algorithms, presents issues related to their usefulness in the context of sustainable development. Previous research related to the essence and research directions of bee algorithms, the origins and basic assumptions of these algorithms qualified as collective intelligence, and actually relating to concepts of swarm algorithms known from nature. The main part concerns the application areas and types of tasks that can be supported by such approaches. We can observe the continuous development of the original concepts of the discussed algorithms, directions of further research will be included in the summary of the article.
Mieczysław L. Owoc
Automatic Coral Morphotypes Detection with YOLO: A Deep Learning Approach for Efficient and Accurate Coral Reef Monitoring
Abstract
Coral reefs harbor a large portion of marine biodiversity but are declining rapidly. Conservation efforts rely on monitoring coral abundance and composition as predominant indicators of ecosystem health and management success. However, manual monitoring of coral abundance proves arduous where artificial intelligence-based automatization can help improve efficiency and accuracy. This paper presents a methodology using YOLOv5-based deep learning for automatic detection of corals and classification by morphotype, representing an important step toward streamlining machine-assisted coral reef monitoring. The research addresses the escalating need for precise and timely ecosystem assessments amidst increasing ecological shifts of coral reefs. Using state-of-the-art object detection techniques, the study strives to streamline the detection and classification of diverse coral morphotypes, which is essential for understanding reef dynamics and assessing conservation efforts. To train and evaluate our system, we use a dataset consisting of 280 original underwater coral reef images. We increased the number of annotated images to 388 by manipulating images using data augmentation techniques, which can improve model performance by providing more diverse examples for training. Our system leverages the YOLOv5 algorithm’s real-time object detection capabilities, enabling efficient and accurate coral detection. We used YOLOv5 to extract discriminating features from the annotated dataset, enabling the system to generalize, including previously unseen underwater images. The successful implementation of the automatic coral morphotypes detection system with YOLOv5 on our original image dataset highlights the potential of advanced computer vision techniques for coral reef research and conservation.
Younes Ouassine, Jihad Zahir, Noël Conruyt, Mohsen Kayal, Philippe A. Martin, Eric Chenin, Lionel Bigot, Regine Vignes Lebbe
Answering Key Questions About Air Pollution Dynamics in Ho Chi Minh City: A Spatiotemporal Analysis Using the XAI-SHAP Clustering Approach
Abstract
This study investigates the impact of various air pollutants in Ho Chi Minh City (HCMC), Vietnam, employing a supervised clustering approach that integrates Shapley Additive exPlanations (SHAP) values with a publicly available outdoor air quality dataset. The research focuses on identifying influential pollutant factors, affected city regions, their impact on surrounding areas, and the effectiveness of supervised clustering using SHAP values in pinpointing intersections between impacted stations. A feature set from Rakholia et al. (2023) is utilized in conjunction with tree-based machine learning models for each target time series. This approach provides insights into the contribution of different variables through the lens of eXplainable artificial intelligence (XAI)-SHAP methodology. The study also explores intersections between affected areas using the Uniform Manifold Approximation and Projection technique. Key findings indicate a pronounced influence of Nitrogen Dioxide (NO2) and Temperature on morning air quality, largely attributed to traffic congestion. Conversely, Carbon Monoxide (CO) and Ozone (O3) significantly affect the afternoon air quality, influenced by solar radiation and industrial activities. Spatial and temporal analysis of air pollution variables suggests that mapping interference between monitoring stations can enhance categorization accuracy in certain city’s parts. This research provides valuable insights that can significantly benefit urban planning, time-sensitive policy development, and targeted intervention strategies. These findings are intended to support policymakers and stakeholders in addressing air pollution in HCMC more effectively.
Polat Goktas, Rajnish Rakholia, Ricardo S. Carbajo
Crowdsourcing as a Tool for Smart City Within Sustainable Development
Abstract
Increasingly, more individuals are opting to reside in urban areas, a trend that is driving city administrators to develop innovative strategies for effective urban resource management. The Smart City (SC) concept, increasingly adopted in various global metropolises, significantly enhances residents’ quality of life, safety, social participation, and operational efficiency of urban areas, aligning with the crucial paradigm of Sustainable Development (SD). A key initiative underpinning the SC model is crowdsourcing, a method whereby urban administrations gather knowledge and feedback from their citizens. This approach, increasingly focused on SD aspects, is gaining prominence due to the current climate challenges and rapid dynamics of urban development.
This paper presents findings from an examination of scientific literature indexed in the Scopus database, using keywords: ‘crowdsourcing’, ‘smart cities’, and ‘sustainable development’. Employing Classical Literature Review (CLR) and Systematic Literature Review (SLR) methodologies, supplemented by bibliometric analysis with VOSviewer software, this study conducts an in-depth investigation of keywords and research themes closely linked to the development of crowdsourcing within the SC framework, specifically in the context of SD. The results are presented in the form of bibliometric maps and tables, highlighting identified keywords.
SCs integrate science and technology to enhance urban life, addressing contemporary challenges by transforming urban areas into intelligent, efficient, and sustainable spaces. Crowdsourcing plays a pivotal role in this transformation, enabling community involvement in urban planning and decision-making. This study on crowdsourcing as a tool for SC initiatives within SD reveals numerous practical and theoretical implications and opens avenues for future research.
Lukasz Przysucha, Adam Sulich
Model of Relationship Between Circular Economy and Industry 5.0
Abstract
This research examines the interaction between Industry 5.0's role in developed European nations and their efforts to achieve a Circular Economy (CE). Both concepts are innovative socially, sustainable ecologically, and viable economically. Specifically, the study focuses on the European Union (EU), exploring how CE initiatives are integrated within its member states using various representational variables. The primary goal of this research is to develop an econometric model that elucidates the impact of CE activities on the advancement toward Industry 5.0, potentially indicating a significant shift in the approach to technology and the environment. Within the EU context, the study delves into the complexities of CE, utilizing selected indicators and data from Eurostat. Methodologically, the research employs both the taxonomic method and regression analysis. It also highlights the importance of Industry 5.0 in incorporating Sustainable Development (SD) principles across various sectors, thus underlining its key role in driving broader economic transformation through technological innovation. The study posits that the path of economic development is influenced by strategic management where the outcomes of Industry 5.0, in concert with CE principles, play a pivotal role in promoting green jobs, bolstered by eco-innovation and green patents.
Adam Sulich
Reduction of Carbon Dioxide Emissions of IT Hardware
Abstract
Every software needs hardware to be run on. Nowadays we encounter urgent need to mitigate the adverse effects of climate change. It prompted a paradigm shift towards the development of low carbon dioxide (CO2) emission hardware infrastructure. This scientific paper presents a thorough analysis of the current state of low CO2 emission hardware infrastructure, highlighting its significance in achieving sustainable and environmentally friendly technological advancements. The paper begins by highlighting the global greenhouse gas emissions and their significance in changing the climate. It emphasizes the crucial role of hardware infrastructure in this context, as the energy consumption and carbon footprint of data centers, communication networks, and other hardware-intensive systems continue to rise. Next, in the paper have been analyzed various strategies and technologies that have been developed to reduce CO2 emissions during computations. These include energy efficient designs, advanced cooling techniques, renewable energy integration, audits and controls, and optimization algorithms such modern AI tools. The advantages and limitations of each approach are discussed, with a focus on their potential for widespread adoption and scalability.
The paper concludes by outlining the future prospects and challenges associated with low CO2 emission hardware infrastructure. It emphasizes the need for continued research and innovation to overcome existing barriers and accelerate the adoption of environmentally friendly hardware systems on a global scale.
The research methods used in the article are literature analysis and analysis of commercial research reports.
Kamil Hudaszek, Iwona Chomiak-Orsa, Saeed Abdullah M. AL-Dobai
Backmatter
Metadaten
Titel
Artificial Intelligence for Knowledge Management, Energy and Sustainability
herausgegeben von
Eunika Mercier-Laurent
Gülgün Kayakutlu
Mieczyslaw Lech Owoc
Abdul Wahid
Karl Mason
Copyright-Jahr
2024
Electronic ISBN
978-3-031-61069-1
Print ISBN
978-3-031-61068-4
DOI
https://doi.org/10.1007/978-3-031-61069-1