Wyniki wyszukiwania - BazTech

1

Co w sep piszczy: czyli oddział elektroniki, informatyki, telekomunikacji stowarzyszenia elektryków polskich

Romaniuk Ryszard

Elektronika : konstrukcje, technologie, zastosowania

|

2026

|

Vol. 67, Nr 2

22--25

PL

W artykule autor przygląda się możliwościom działania i rozwoju Oddziału Elektroniki, Informacji, Telekomunikacji SEP w nowej kadencji obejmującej lata 2026-2030. Oddział EIT SEP będący komponentem dużego stowarzyszenia naukowo-technicznego, posiadający samodzielność gospodarczą, musi dostosowywać się aktywnie do zmiennych warunków ekonomicznych i społecznych, identyfikując obszary społeczne największej potrzeby i przydatności. Wiele z tych obszarów ma charakter klasyczny, jak kształcenie ustawiczne inżynierów, kursy i szkolenia, wsparcie dla firm, współpraca ze środowiskiem naukowym i gospodarczym, doradztwo i rzeczoznawstwo techniczne, zawodowa działalność środowiskowa i społeczna. Te obszary i metody działania podlegają zmianom, pojawiają się nowe, jak uczenie na odległość, źródłowe bazy wiedzy, a przede wszystkim sztuczna inteligencja. Jak w takich warunkach ma odnaleźć się O. EIT SEP, aby być pożytecznym i oferować dobre i nowoczesne usługi zawodowe, a także przeżyć gospodarczo, jako organizacja mieszana - stowarzyszenie wyższej użyteczności publicznej, ale i specyficzna, niedochodowa jednostka gospodarcza?

EN

In this article, the author examine the operational and development opportunities for the Electronics, Information, and Telecommunications Branch of the SEP (SEP) in its new term of office, spanning the period 2026- 2030. As a component of a large scientific and technical association, the EIT SEP branch, with its economic independence, must actively adapt to changing economic and social conditions. In these changing circumstances, we strive to identify the social areas of greatest need and utility for the activities of a scientific and technical association. Many of these areas are traditional in nature, such as continuing education for engineers, courses and training, support and recommendations for companies, collaboration with science and industry, technical consulting and expertise, and professional environmental and social activities. These areas and their methods are subject to change, with new areas emerging, such as distance learning, knowledge bases, and, above all, artificial intelligence. How can the O. EIT SEP navigate these conditions to be useful, relevant, offer high-quality, modern professional services, and also survive economically as a mixed organization - an association of higher public utility and an special, non-profit economic entity?

2

ML-based models in support of selected legal and administrative activities

Łaganowska Monika, Mikołajewska Maria, Mikołajewski Dariusz

Studia i Materiały Informatyki Stosowanej

|

2025

|

T. 17, nr 2

42--47

EN

This article discusses the application of machine learning (ML) models in improving legal and administrative processes. It highlights how ML techniques such as natural language processing and predictive analytics can automate routine tasks such as document classification, legal research, and case outcome prediction. The authors discuss the benefits of ML-based systems, including increased efficiency, reduced human error, and increased access to justice. Ethical issues are addressed, particularly regarding algorithmic bias, transparency, and accountability in decision-making. Case studies are presented to illustrate the real-world implementation of these technologies in courts and public administration. The article concludes by emphasizing the need for interdisciplinary collaboration and regulatory frameworks to ensure responsible and effective integration of ML in legal domains.

PL

Artykuł omawia zastosowanie modeli uczenia maszynowego (ML) w ulepszaniu procesów prawnych i administracyjnych. Podkreśla, w jaki sposób techniki ML, takie jak przetwarzanie języka naturalnego i analityka predykcyjna, mogą automatyzować rutynowe zadania, takie jak klasyfikacja dokumentów, badania prawne i przewidywanie wyników spraw. Autorzy omawiają korzyści płynące z systemów opartych na ML, w tym zwiększoną wydajność, zmniejszenie liczby błędów ludzkich i zwiększony dostęp do wymiaru sprawiedliwości. Poruszane są kwestie etyczne, w szczególności dotyczące stronniczości algorytmicznej, przejrzystości i odpowiedzialności w podejmowaniu decyzji. Przedstawiono studia przypadków, aby zilustrować rzeczywiste wdrożenie tych technologii w sądach i administracji publicznej. Artykuł kończy się podkreśleniem potrzeby interdyscyplinarnej współpracy i ram regulacyjnych w celu zapewnienia odpowiedzialnej i skutecznej integracji ML w domenach prawnych.

3

Generative AI in support of selected legal and administrative activities

Mikołajewska Maria, Łaganowska Monika, Mikołajewski Dariusz

Studia i Materiały Informatyki Stosowanej

|

2025

|

T. 17, nr 1

66--71

EN

Generative AI (Gen AI) transforms legal and administrative work by helpingto rapidly draft contracts, pleadings, and routine correspondence, freeing professionals’ time to focus on more demanding tasks, such as valuable analysis and strategy.Accelerates legal research through natural language queries and summaries, revealing precedents and regulations that match nuanced fact patterns in seconds.In administrative contexts, generative models automate form generation, policy templates, and multilingual communication, reducing administrative errors and turnaround times.When combined with ingest-enhanced generation and audit trails, these systems enable transparent sourcing, version control, and compliance monitoring, meeting evidentiary and procedural requirements.The result is a hybrid workflow where human expertise guides judgmental decisions while AI enables scalable, cost-effective document development, research, and management.Słowa kluczowe: Computer science, artificial intelligence, generative AI, legal applications, administrative applications.

PL

Generatywna sztuczna inteligencja (GenAI) zmienia oblicze pracy prawnej i administracyjnej, pomagając szybko opracować umowy, pisma procesowe i rutynową korespondencję, uwalniając czas profesjonalistów, aby mogli skupić się na bardziej wymagających zadaniach. np. wartościowej analizie i strategii. Przyspiesza badania prawne poprzez zapytania i podsumowania w języku naturalnym, ujawniając precedensy i przepisy, które pasują do niuansów wzorców faktów w ciągu kilku sekund. W kontekstach administracyjnych modele generatywne automatyzują generowanie formularzy, szablony zasad i komunikację wielojęzyczną, redukując błędy administracyjne i czas realizacji. W połączeniu z generowaniem rozszerzonym o pobieranie i elementami audytu systemy te umożliwiają przejrzyste pozyskiwanie, kontrolę wersji i monitorowanie zgodności, spełniając wymogi dowodowe i proceduralne. Rezultatem jest hybrydowy przepływ pracy, w którym ludzka wiedza specjalistyczna kieruje decyzjami wymagającymi osądu, podczas gdy sztuczna inteligencja zapewnia skalowalne, ekonomiczne opracowywanie, badania i zarządzanie dokumentacją.

4

IT challenges of speech therapy and physiotherapy of the oral-facial tract

Mikołajewska Emilia

Studia i Materiały Informatyki Stosowanej

|

2024

|

T. 16, nr 2

28--31

EN

Speech therapy and physiotherapy of the oro-facial tract can already be effectively supported by IT solutions, both in the area of diagnosis and therapy. However, this requires an integrated approach based on interdisciplinary cooperation between speech therapists, physiotherapists and computer scientists. The aim of the article is to assess the extent to which current IT capabilities are used to support diagnostics and therapy in speech therapy and physiotherapy in the oro-facial tract and to determine the potential for their further stimulated development.

PL

Logopedię i fizjoterapię układu ustno-twarzowego można już skutecznie wspierać rozwiązaniami informatycznymi, zarówno w obszarze diagnostyki, jak i terapii. Wymaga to jednak zintegrowanego podejścia, opartego na interdyscyplinarnej współpracy logopedów, fizjoterapeutów i informatyków. Celem artykułu jest ocena, w jakim stopniu obecne możliwości informatyki są wykorzystywan dla wsparcia diragnostyki i terapii w logopedii i fizjoterapii traktuv oro-facjalnego oraz określenie potencjału do ich dalszego stymulowanego rozwoju.

5

Kompetencje informatyczne uczniów kształtowane w trakcie edukacji podstawowej i ponadpodstawowej na podstawie oceny „krzyżowej” – nauczycieli i uczniów szkół średnich. Województwo dolnośląskie

Wiśniewska-Paź Barbara, Trojanowski Paweł

Cybersecurity and Law

|

2024

|

Vol. 12 (2)

78--93

PL

Kompetencje informatyczne, tj. nie tylko umiejętność obsługi komputera, lecz także pewien poziom elastyczności w używaniu oprogramowania (świadomość zagrożeń, ograniczeń, ale także możliwości w nim tkwiących) to podstawowy zasób określonych kompetencji społecznych w ztechnologizowanym świecie i jednocześnie jeden z głównych poziomów kompetencji cyfrowych obok informacyjnego i funkcjonalnego, istotny w społeczeństwie określanym mianem informatycznego. Jest on kształtowany zarówno w procesach żywiołowych (indywidualny użytek), jak i w procesie edukacyjnym oraz środowisku pracy. Z punktu widzenia pożądanych efektów – edukacyjny ma istotne znacznie. Czy szkoła polska jest przygotowana pod względem programowym, sprzętowym i kadrowym do tego typu wyzwań? Czy nowy rodzaj wiedzy, jaką ma przekazywać, nie zostanie rozmyty w tradycyjnym stylu nauczania? Czy kultura informatyczna i nowe słownictwo opisujące nieistniejące przed dekadami obszary społecznej praktyki staną się czymś powszechnym czy też raczej wyznaczą nowe kryteria społecznych podziałów? To pytania, które leżą u źródła zainteresowania zagadnieniem przedstawionym w niniejszym artykule, w którym zostały zaprezentowane wyniki badań przeprowadzonych wśród uczniów i nauczycieli szkół po¬nadpodstawowych. W artykule są omawiane kompetencje informatyczne uczniów szkół ponadpodstawowych na podstawie oceny uzyskanej w szkole podstawowej. Zderzenie samoocen uczniowskich z ocenami nauczycieli pozwala zwrócić uwagę na pewne niedoskonałości nauczania informatyki w polskich szkołach powszechnych.

EN

IT competences, not only the ability to operate a computer, but also a certain level of flexibility when using software (awareness of threats, limitations, but also opportunities that lie in it) is a specific social resource. Essential in a society known as information technology. It is shaped both in natural processes (individual use) and in the educational process. From the point of view of the desired effects, the latter is of much greater importance. However, is the Polish school prepared, in terms of programme, equipment and staff, for this type of challenge? Will the new kind of knowledge it is supposed to impart be „watered down” by the old-fashioned way of teaching? Will IT culture and a new range of vocabulary describing areas of social practice that did not exist decades ago become something common, or will they rather set new criteria for social divisions? These are the questions that lie at the source of interest in the issue referred to in this article, which presents a report on research conducted among students and teachers of secondary schools. The article discusses the IT competences of secondary school students in relation to the grade obtained in primary school. The clash of self-assessments of „students” with teachers’ assessments allows to highlight the imperfections of teaching computer science in Polish elementary schools.

6

SrpCNNeL: Serbian Model for Named Entity Linking

Ikonić Nešić Milica, Petalinkar Saša, Stanković Ranka, Utvić Miloš, Kitanović Olivera

Annals of Computer Science and Information Systems

|

2024

|

Vol. 39

465--473

EN

This paper presents the development of a Named Entity Linking (NEL) model to the Wikidata knowledge base for the Serbian language named SrpCNNeL. The model was trained to recognize and link seven different named entity types (persons, locations, organisations, professions, events, demonyms, and works of art) on the dataset containing sentences from novels, legal documents, as also sentences generated from the Wikidata knowledge base and Leximirka lexical database. The resulting model demonstrated robust performance, achieving an F1 score of 0.8 on the test set. Considering that the dataset contains the highest number of locations linked to the knowledge base, an evaluation was conducted on an independent dataset and compared to the baseline Spacy Entity Linker for locations only.

7

Spoken Language Corpora Augmentation with Domain-Specific Voice-Cloned Speech

Czyżnikiewicz Mateusz, Bondaruk Łukasz, Kubiak Jakub, Wiącek Adam, Degórski Łukasz, Kubis Marek, Skórzewski Paweł

Annals of Computer Science and Information Systems

|

2024

|

Vol. 39

579--584

EN

In this paper we study the impact of augmenting spoken language corpora with domain-specific synthetic samples for the purpose of training a speech recognition system. Using both a conventional neural text-to-speech system and a zero-shot one with voice cloning ability we generate speech corpora that vary in the number of voices. We compare speech recognition models trained with addition of different amounts of synthetic data generated using these two methods with a baseline model trained solely on voice recordings. We show that while the quality of voice-cloned dataset is lower, its increased multivoiceity makes it much more effective than the one with only a few voices synthesized with the use of a conventional neural text-to-speech system. Furthermore, our experiments indicate that using low variability synthetic speech quickly leads to saturation in the quality of the ASR whereas high variability speech provides improvement even when increasing total amount of data used for training by 30%.

8

Secretary problem revisited: Optimal selection strategy for top candidates using one try in a generalized version of the problem

Štěpánek Lubomír

Annals of Computer Science and Information Systems

|

2024

|

Vol. 39

719--724

EN

This paper explores a novel variation of the classical secretary problem, commonly referred to as the marriage or best choice problem. In this adaptation, a decision-maker sequentially dates n ∈ N candidates, each uniquely ranked without ties from 1 to n. The decision strategy involves a preliminary non-selection phase of the first d ∈ N candidates where, d < n, following which the decision-maker commits to the first subsequent candidate who surpasses all previously evaluated candidates in quality. The central focus of this study is the derivation and analysis of P (d, n, k), which denotes the probability that the selected candidate, under the prescribed strategy, ranks among the top k ∈ N overall candidates, where k ≤ n. This investigation employs combinatorial probability theory to formulate P (d, n, k) and explores its behavior across various parameter values of d, n, and k. Particularly, we seek to determine in what fraction of the entire decision process should a decision-maker stop the non-selection phase, i.e., we search for the optimal proportion d/n ,that maximizes the probability P (d, n, k), with a special focus on scenarios where k is in generally low. While for k = 1, the problem is simplified to the classical secretary problem with d/n ≈ 1/e , our findings suggest that the strategy’s effectiveness is optimized for portion d/n decreasing below 1/e as k increases. Also, intuitively, probability P (d, n, k) increases as k increases, since the number of acceptable top candidates increases. These results not only extend the classical secretary problem but also provide strategic insights into decision-making processes involving ranked choices, sequential evaluation, and applications of searching not necessarily the best candidate, but one of the best candidates.

9

Reinforcement Learning based Intelligent System for Personalized Exam Schedule

Barone Marco, Ciaschi Matteo, Ullah Zaib, Piccardi Armando

Annals of Computer Science and Information Systems

|

2024

|

Vol. 39

549-–553

EN

Personalized learning has been proving to be useful concept in the learning of a student. Artificial Intelligence (AI) which has revolutionized many aspects of our lives has also been glowingly used in the education sector. One of the fascinating AI technique, the Reinforcement Learning (RL) is considered as the perfect tool to develop personalized solution in the education. RL algorithms have the ability to take into account personal characteristics of each student. This work presents the development of personalized exam scheduler using RL. The intelligent examination scheduler consider several parameters for training such as age, academic year, past education performance, discipline, number of courses, and gap between two exams. The trained RL agent then able to provide examination schedule to a student depending on a student personal record, interests and abilities. The preliminary results are encouraging and more research would bring useful contribution of AI in various aspects of learning process of a student.

10

Pareto Optimal Solutions of the Biobjective Minimum Length Minimum Risk Spanning Trees Problem

Laskov Lasko, Marinov Marin

Annals of Computer Science and Information Systems

|

2024

|

Vol. 39

405--416

EN

We propose an exact method that finds the complete Pareto front of the biobjective minimum length minimum risk spanning trees problem. The proposed method is based on the solution of two subproblems. The first subproblem is to find a list of all minimum spanning trees with respect of the length criterion. The second subproblem is to compose the complete Pareto front itself, based on the list of all Pareto optimal trees with minimal length. We provide detailed mathematical proofs for the correctness and running time complexity of all proposed algorithms. We also illustrate all algorithms with detailed numerical examples.

11

Key Factors Influencing Mobile Banking Adoption in Saudi Arabia

Alzahrani Amal, Beloff Natalia, White Martin

Annals of Computer Science and Information Systems

|

2024

|

Vol. 39

531--536

EN

The introduction of mobile banking has revolutionized traditional financial practices, enhancing efficiency, customer experiences, and business models globally. Despite these advancements, mobile banking adoption remains low in Saudi Arabia. This paper seeks to fill this gap by examining the significance of factors that either drive or hinder adoption. We propose a novel model integrating factors from the DeLone and McLean (D&M) model and factors from the Unified Theory of Acceptance and Use of Technology (UTAUT2) model, complemented by additional factors. Quantitative data was collected through online questionnaire from a diverse sample of Saudi banking customers, supplemented by qualitative insights from customer interviews. Findings revealed that net benefits, compatibility, facilitating conditions, and trust positively influence adoption, while literacy levels and digital skills pose barriers. Our study offers a significant theoretical contribution by synthesizing multiple models and enriches understanding of mobile banking adoption, aiding future research and industry decision making.

12

Impact of Spelling and Editing Correctness on Detection of LLM-Generated Emails

Gryka Paweł, Gradoń Kacper, Kozłowski Marek, Kutyła Miłosz, Janicki Artur

Annals of Computer Science and Information Systems

|

2024

|

Vol. 39

603--608

EN

In this paper, we investigated the impact of spelling and editing correctness on the accuracy of detection if an email was written by a human or if it was generated by a language model. As a dataset, we used a combination of publicly available email datasets with our in-house data, with over 10k emails in total. Then, we generated their “copies'' using large language models (LLMs) with specific prompts. As a classifier, we used random forest, which yielded the best results in previous experiments. For English emails, we found a slight decrease in evaluation metrics if error-related features were excluded. However, for the Polish emails, the differences were more significant, indicating a decline in prediction quality by around 2% relative. The results suggest that the proposed detection method can be equally effective for English even if spelling- and grammar-checking tools are used. As for Polish, to compensate for error-related features, additional measures have to be undertaken.

13

Impact of Local Geometry on Methods for Constructing Protein Conformations

Da Rocha W., Malliavin T. E., Mucherino A., Liberti L.

Annals of Computer Science and Information Systems

|

2024

|

Vol. 39

677--681

EN

The prediction of protein structures is an important problem in molecular biology. In spite of the large efforts from the research community, and of the recent development of artificial intelligence tools specifically designed for this problem, a complete and definitive solution to the problem has not been found yet. This work is based on the observation that many tools for the prediction of protein conformations rely on both local and non-local geometrical information, even though the non-local information can be very hard to identify within the desired precision in some particular situations. For this reason, we explore in this work the effect of local geometry on methods capable of constructing protein conformations. This initial study has the final aim of devising new alternative methods where the predictions may be guided mainly by the local geometry of proteins.

14

Evolving the Enterprise Software Systems Landscape: Towards Anti-Patterns in Smalltalk-to-Java Code Transformation

Bělohoubek Marek, Pergl Robert

Annals of Computer Science and Information Systems

|

2024

|

Vol. 39

567--572

EN

In the rapidly evolving landscape of enterprise software systems, there is a marked escalation in the proliferation of new technologies, tools, languages, and methodologies daily. These innovations are pivotal not only for the development of new systems but also for the maintenance and augmentation of existing infrastructures. Consequently, it is imperative to devise systems that are responsive to these advancements, fostering the integration of novel tools and methodologies into the current systems. This integration often necessitates mechanisms for transforming source code across diverse programming languages. In the course of developing a transformation tool from Smalltalk to Java, we encountered several code patterns that significantly impede the transformation process. This paper aims to elucidate one such transformation anti-pattern. We provide a comprehensive overview, a formal delineation, illustrations derived from actual code, and propose refactoring strategies for both Smalltalk and Java environments.

15

Efficient Maritime Healthcare Resource Allocation Using Reinforcement Learning

Hasan Tehreem, Batool Farwa, Fiorino Mario, Tretola Giancarlo, Abbas Musarat

Annals of Computer Science and Information Systems

|

2024

|

Vol. 39

615--620

EN

The allocation of healthcare resources on ships is crucial for safety and well-being due to limited access to external aid. Proficient medical staff on board provide a mobile healthcare facility, offering a range of services from first aid to complex procedures. This paper presents a system model utilizing Reinforcement Learning (RL) to optimize doctor-patient assignments and resource allocation in maritime settings. The RL approach focuses on dynamic, sequential decision-making, employing Q-learning to adapt to changing conditions and maximize cumulative rewards. Our experimental setup involves a simulated healthcare environment with variable patient conditions and doctor availability, operating within a 24-hour cycle. The Q-learning algorithm iteratively learns optimal strategies to enhance resource utilization and patient outcomes, prioritizing emergency cases while balancing the availability of medical staff. The results highlight the potential of RL in improving healthcare delivery on ships, demonstrating the system's effectiveness in dynamic, time-constrained scenarios and contributing to overall maritime safety and operational resilience.

16

Dynamic relationship between population densities and air quality in the four largest Norwegian cities

Zhivkov Petar, Kesarovski Todor

Annals of Computer Science and Information Systems

|

2024

|

Vol. 39

713--718

EN

Air pollution is a significant cause of health problems and disease worldwide. Considering the rapid urbanisation at a global scale in recent decades, resulting in more and more people in urban areas, cities deserve special attention in this regard. In this paper, we use air quality measurement data from 2010 to 2023 in the four largest Norwegian cities (Oslo, Bergen, Trondheim and Stavanger) and correlate it with the evolution of population densities for the same period. The empirical analysis focuses on nitrogen dioxides (NO2) and particular matter (PM2.5 and PM10) as critical pollutants in urban areas to verify whether their concentrations are affected by the increase in population densities for individual municipalities. In addition, we also correlate the data on air pollutants with different natural indicators such as temperature, air pressure, humidity and wind and the rate of motorisation in the cities of interest.

17

Disease Diagnosis On Ships Using Hierarchical Reinforcement Learning

Batool Farwa, Hasan Tehreem, Tretola Giancarlo, Ullah Zaib, Abbas Musarat

Annals of Computer Science and Information Systems

|

2024

|

Vol. 39

555-–559

EN

Every year about 30 million people travel by ships world wide often in extreme weather conditions and also in polluted environment due to ship's fuel combustion and many other factors that impacts the health of both passengers and crew staff so there is a need of medical staff but that's not always available so we introduce an a model based on Reinforcement learning(RL) that is used as the key approach in dialogue system.We incorporates Hierarchical reinforcement learning (HRL) model with the layers of Deep Q-Network for dialogue oriented diagnosis system.policy learning is integrated as policy gradients are already defined.We created two stage hierarchical strategy.We used the hierarchical structure with double layer policies for automatic disease diagnosis.Double layer means it splits the task into sub-tasks named as high-state strategy and low level strategy.It has a component called user simulator that communicates with patient for symptom collection low level agent inquire symptoms.Once its done collecting it sends results to high level agent which activates the D-classifier for last diagnosis.When its done its send back by user simulator to patients to verify diagnosis made.Every single diagnosis made has its own reward that trains the system.

18

Comparing Lazy Constraint Selection Strategies in Train Routing with Moving Block Control

Engels Stefan, Wille Robert

Annals of Computer Science and Information Systems

|

2024

|

Vol. 39

585--590

EN

Railroad transportation plays a vital role in the future of sustainable mobility. Besides building new infrastructure, capacity can be improved by modern train control systems, e.g., based on moving blocks. At the same time, there is only limited work on how to optimally route trains using the potential gained by these systems. Recently, an initial approach for train routing with moving block control has been proposed to address this demand. However, detailed evaluations on so-called lazy constraints are missing, and no publicly available implementation exists. In this work, we close this gap by providing an extended approach as well as a flexible open-source implementation that can use different solving strategies. Using that, we experimentally evaluate what choices should be made when implementing a lazy constraint approach. The corresponding implementation and benchmarks are publicly available as part of the Munich Train Control Toolkit (MTCT) at https://github.com/cda-tum/mtct.

19

An Improved Genetic Algorithm for Set Cover using Rosenthal Potential

Tayebi Dena, Ray Saurabh, Ajwani Deepak

Annals of Computer Science and Information Systems

|

2024

|

Vol. 39

689--694

EN

A major issue with heuristics for set-cover problem is that they tend to get stuck in a local optimum typically because a large local move is necessary to find a better solution. A recent theoretical result shows that replacing the objective function by a proxy (which happens to be Rosenthal potential function) allows escaping such local optima even with small local moves albeit at the cost of an approximation factor. The Rosenthal potential function thus has the effect of smoothing the optimization landscape appropriately so that local search works. In this paper, we use this theoretical insight to design a simple but robust genetic algorithm for weighted set cover. We modify the fitness function as well as the crossover operator of the genetic algorithm to leverage the Rosenthal potential function. We show empirically this greatly improves the quality of the solutions obtained especially in examples where large local moves are required. Our results are better than existing state of the art genetic algorithms and also comparable in performance with the recent local search algorithm NuSC (carefully engineered for set cover) on benchmark instances. Our algorithm, however, performs better than NuSC on simple synthetic instances where starting from an initial solution, large local moves are necessary to find a solution that is close to optimal. For such instances, our algorithm is able to find near optimal solutions whereas NuSC either takes a very long time or returns a much worse solution.

20

A network clustering method based on intersection of random spanning trees

Hajdu László, London András, Pluhár András

Annals of Computer Science and Information Systems

|

2024

|

Vol. 39

609--614

EN

We use a special edge centrality measure for node clustering in complex networks. The measure is based on the `spanning tree intersection' value motivated by previous work on the intersection and minimum expected overlap of random spanning trees in complex networks. First, we show that this new metric differs from some well-known edge centralities on random network models and real-world networks. Then, we show the applicability of the metric for clustering the nodes and point out some advantages over some other edge centrality based hierarchical clustering methods.