Search

This study identifies common product combinations within the grocery sector across store formats. Although store formats and product combinations have been studied previously, they have been examined separately in different contexts and not together. This constitutes a research gap that this study aims to address. The objective is to propose strategies that improve the product offering for each store format and its target customer. The combinations are developed at product category level as well as on item level, where the combinations are formed for one product group on item level. Additionally, the study examines whether the most associated categories contribute to overall profitability compared to categories with the highest sales and receipt share. The results provide insights into which assortment strategies companies should focus on to better reach their customers. Customer receipts from Coop Sweden are used, with data from Coop, Stora Coop, and X:-tra. Market basket analysis, specifically the apriori algorithm, is employed to create- the associations. The results showed the most associated categories and items, and the specific combinations. At the item level, the category cheese was chosen due to its high sales and association. The study’s findings suggest that Coop should adapt its offering with high-quality and environmentally friendly products for sporadic and environmentally conscious customers. Stora Coop should offer larger packages for planned purchases and tailor its assortment for breakfast and dinner planning. X:- tra stores should focus on cost-conscious, value-based baskets and clearly defined baskets, where their customers often buy private labels and have well-defined shopping patterns., Denna studie identifierar frekventa produktkombinationer inom dagligvaruhandeln för olika butiksformat. Trots att butiksformat och produktkombinationer har studerats tidigare, har de undersökts i olika sammanhang och inte tillsammans. Detta utgör ett forskningsgap som denna studie avser att fylla. Målet är att föreslå strategier som förbättrar produktutbudet för varje butiksformat och dess kundgrupper. Kombinationerna genereras både på produktkategorinivå och på artikelnivå, där alla kombinationer på artikelnivå produceras från en produktgrupp. Dessutom undersöks om de mest associerade kategorierna bidrar till den totala lönsamheten jämfört med kategorier med högst försäljning och högst kvittoandel. Resultaten ger insikter om vilka sortimentsstrategier företag bör fokusera på för att bättre nå sina kunder. Kundkvitton från Coop Sverige används, med data från Coop, Stora Coop och X:- tra. Marknadskorgsanalys, specifikt apriori-algoritmen, för att skapa associationer mellan butikstyperna. Resultaten visade de mest associerade produktkategorierna och artiklarna samt signifikanta kombinationer. På artikelnivå valdes kategorin ost då kategorin visade på hög försäljning och association. Studiens resultat tyder på att Coop bör anpassa sitt sortiment med kvalitativa och miljövänliga produkter för sporadiska och miljömedvetna kunder. Stora Coop bör erbjuda större förpackningar för planerade inköp och anpassa sitt sortiment för frukost- och middagsplanering. X:-tra-butiker bör fokusera på kostnadsmedvetna, värdebaserade korgar samt tydligt definierade varukorgar, där deras kunder ofta köper egna varumärken och har väldefinierade inköpsmönster.

Sarcomas are rare mesodermal tumors of heterogeneous nature and have a higher incidence in children. The relative 5-year survival rate for patients with metastatic sarcoma is usually low. Standard treatment for sarcomas involves surgical resection, and investigating the genetic basis of these tumors through genome-wide analysis is crucial due to their rarity and late diagnosis. This work proposes a methodology that combines preprocessing, feature selection and association rule mining to identify relevant genes and significant relationships in biological data from sarcoma patients. Our study aims to identify the relationships between metastasis-associated genes and patient survival of less than 5 years. The proposed approach was applied to a sarcoma dataset containing data on gene expression, metastasis occurrence, and survival time, revealing a set of biologically relevant gene interactions associated with sarcoma metastasis and low survival rates. The combined use of these techniques can facilitate the identification of biomarkers or gene signatures associated with the disease and provide insight into the underlying biological mechanisms involved in sarcomas.

Background: Association rules are more comprehensive and understandable than fault-prone module predictors (such as logistic regression model, random forest and support vector machine). One of the challenges is that there are usually too many similar rules to be extracted by the rule mining. Aim: This paper proposes a rule reduction technique that can eliminate complex (long) and/or similar rules without sacrificing the prediction performance as much as possible. Method: The notion of the method is to removing long and similar rules unless their confidence level as a heuristic is high enough than shorter rules. For example, it starts with selecting rules with shortest length (length=1), and then it continues through the 2nd shortest rules selection (length=2) based on the current confidence level, this process is repeated on the selection for longer rules until no rules are worth included. Result: An empirical experiment has been conducted with the Mylyn and Eclipse PDE datasets. The result of the Mylyn dataset showed the proposed method was able to reduce the number of rules from 1347 down to 13, while the delta of the prediction performance was only. 015 (from. 757 down to. 742) in terms of the F1 prediction criteria. In the experiment with Eclipsed PDE dataset, the proposed method reduced the number of rules from 398 to 12, while the prediction performance even improved (from. 426 to. 441.) Conclusion: The novel technique introduced resolves the rule explosion problem in association rule mining for software proneness prediction, which is significant and provides better understanding of the causes of faulty modules., 2012 19th Asia-Pacific Software Engineering Conference, 4-7 Dec. 2012, Hong Kong, China

This paper describes an empirical study to reveal rules associated with defect correction effort. We defined defect correction effort as a quantitative (ratio scale) variable, and extended conventional (nominal scale based) association rule mining to directly handle such quantitative variables. An extended rule describes the statistical characteristic of a ratio or interval scale variable in the consequent part of the rule by its mean value and standard deviation so that conditions producing distinctive statistics can be discovered As an analysis target, we collected various attributes of about 1,200 defects found in a typical medium-scale, multi-vendor (distance development) information system development project in Japan. Our findings based on extracted rules include: (l)Defects detected in coding/unit testing were easily corrected (less than 7% of mean effort) when they are related to data output or validation of input data. (2)Nevertheless, they sometimes required much more effort (lift of standard deviation was 5.845) in case of low reproducibility, (i)Defects introduced in coding/unit testing often required large correction effort (mean was 12.596 staff-hours and standard deviation was 25.716) when they were related to data handing. From these findings, we confirmed that we need to pay attention to types of defects having large mean effort as well as those having large standard deviation of effort since such defects sometimes cause excess effort., MSR'07:ICSE Workshops 2007 : Fourth International Workshop on Mining Software Repositories, 20-26 May 2007, Minneapolis, MN, USA

Data mining is the process of discovering previously unknown, yet useful, information from data. Data is often large, requiring automated forms of analysis to extract interesting patterns and relationships. To add to this problem, interesting information may be lost when noise is introduced into data. Existing approaches apply lenience to support to accommodate the impacts of noise on frequent patterns. Whilst past research provides improvements to mining within noisy datasets, limited or no work has considered support lenience in relation to rare pattern mining. Rare patterns, as by their name, do not commonly occur. Discovering rare patterns has a variety of applications in areas such as medicine, market analysis, and outlier detection, where commonly occurring events are often uninteresting as they are already well-known and thus do not provide novel insight. I introduce the mining of approximately rare itemsets to discover rare patterns with data in the presence of noise. To mine approximately rare itemsets, an ARI-growth algorithm is proposed. Computational results show ARI-growth is an order of magnitude faster than an Apriori approach. Memory consumption shows similar trends, although, is outperformed by Apriori in one tested case. Approximation with ARI-growth is applicable for rare core pattern recovery.

This paper presents an item selection method in CAT (Computerized Adaptive Testing) based on association rules. According to the correct (1) or incorrect (0) answer of the examinee to each item, association rules are searched with the answered item and its result (1 or 0) as antecedent, and items with greater or lower complexity level than the responded item as a consequent, respectively. The election of the item is performed considering rules with the highest confidence and support. Finally, we include an example to demonstrate the efficacy of the proposed method., En el presente trabajo se propone un método de selección de ítems en los CAT (Computerized Adaptive Testing, Exámenes Adaptativos Computarizados) basado en reglas de asociación. De acuerdo con la respuesta correcta (1) o incorrecta (0) del evaluado a cada pregunta, se buscan reglas de asociación que tengan como antecedente la pregunta contestada y su resultado (1 o 0), y como consecuente, ítems con mayor o menor nivel de complejidad que el ítem respondido, respectivamente. La elección del ítem se realiza considerando las reglas con la mayor confianza y soporte. Finalmente, se incluye un ejemplo para demostrar la eficacia del método propuesto

Machine learning and deep learning have become the most useful and powerful tools in the last years to mine information from large datasets. Despite the successful application to many research fields, it is widely known that some of these solutions based on artificial intelligence are considered black-box models, meaning that most experts find difficult to explain and interpret the models and why they generate such outputs. In this context, explainable artificial intelligence is emerging with the aim of providing black-box models with sufficient interpretability. Thus, models could be easily understood and further applied. This work proposes a novel method to explain black-box models, by using numeric association rules to explain and interpret multi-step time series forecasting models. Thus, a multi-objective algorithm is used to discover quantitative association rules from the target model. Then, visual explanation techniques are applied to make the rules more interpretable. Data from Spanish electricity energy consumption has been used to assess the suitability of the proposal.

Targeting the challenges in the risk analysis of laboratory fire accidents, particularly considering fire accidents in Chinese universities, an integrated approach is proposed with the combination of association rule learning, a Bayesian network (BN), and fuzzy set theory in this study. The proposed approach has the main advantages of deriving conditional probabilities of BN nodes based on historical accident data and association rules (ARs) and making good use of expert elicitation by using an augmented fuzzy set method. In the proposed approach, prior probabilities of the cause nodes are determined based on expert elicitation with the help of an augmented fuzzy set method. The augmented fuzzy set method enables the effective aggregation of expert opinions and helps to reduce subjective bias in expert elicitations. Additionally, an AR algorithm is applied to determine the probabilistic dependency between the BN nodes based on the historical accident data of Chinese universities and further derive conditional probability tables. Finally, the developed fuzzy Bayesian network (FBN) model was employed to identify critical causal factors with respect to laboratory fire accidents in Chinese universities. The obtained results show that H4 (bad safety awareness), O1 (improper storage of hazardous chemicals), E1 (environment with hazardous materials), and M4 (inadequate safety checks) are the four most critical factors inducing laboratory fire accidents., Safety and Security Science

We propose three procedures based on association rules (AR) learning and random forests (RF) to support the specification of a portfolio choice model applied in data from complex choice experiment data, specifically a Participatory Value Evaluation (PVE) choice experiment. In a PVE choice experiment, respondents choose a combination of alternatives, subject to a resource constraint. We combine a methodological-iterative (MI) procedure with AR learning and RF models to support the specification of parameters of a portfolio choice model. Additionally, we use RF model predictions to contrast the validity of the behavioural assumptions of different specifications of the portfolio choice model. We use data of a PVE choice experiment conducted to elicit the preferences of Dutch citizens for lifting COVID-19 measures. Our results show model fit and interpretation improvements in the portfolio choice model, compared with conventional model specifications. Additionally, we provide guidelines on the use of outcomes from AR learning and RF models from a choice modelling perspective., Transport and Logistics, Industrial Design Engineering

This paper proposes a novel approach that combines an association rule algorithm with a deep learning model to enhance the interpretability of prediction outcomes. The study aims to gain insights into the patterns that were learned correctly or incorrectly by the model. To identify these scenarios, an association rule algorithm is applied to extract the patterns learned by the deep learning model. The rules are then analyzed and classified based on specific metrics to draw conclusions about the behavior of the model. We applied this approach to a well-known dataset in various scenarios, such as underfitting and overfitting. The results demonstrate that the combination of the two techniques is highly effective in identifying the patterns learned by the model and analyzing its performance in different scenarios, through error analysis. We suggest that this methodology can enhance the transparency and interpretability of black-box models, thus improving their reliability for real-world applications.

The quality of university teaching is essential for the success of students and the academic excellence of an educational institution. The purpose of this work is to provide a methodology based on the Association Rule technique using the Apriori algorithm to analyze the results obtained from the student evaluation process regarding their satisfaction with the teaching received. This methodology has been applied in programming courses of students of several courses both in the Computer Engineering and Health Engineering degrees at University of Seville, Spain. The proposed methodology can serve as a starting point for a self-improvement process that clearly identifies strengths and weaknesses.

The paper focuses on investigating the impact of artificial agent (machine translator) on human agent (posteditor) using a proposed methodology, which is based on language complexity measures, POS tags, frequent tagsets, association rules, and their summarization. We examine this impact from the point of view of language complexity in terms of word and sentence structure. By the proposed methodology, we analyzed 24 733 tags of English to Slovak translations of technical texts, corresponding to the output of two MT systems (Google Translate and the European Commission’s MT tool). We used both manual (adequacy and fluency) and semiautomatic (HTER metric) MT evaluation measures as the criteria for validity. We show that the proposed methodology is valid based on the evaluation of frequent tagsets and rules of MT outputs produced by Google Translate or of the European Commission’s MT tool, and both postedited MT (PEMT) outputs using baseline methods. Our results have also shown that PEMT output produced by Google Translate is characterized by more frequent tagsets such as verbs in the infinitive with modal verbs compared to its MT output, which is characterized by masculine, inanimate nouns in locative of singular. In the MT output, produced by the European Commission’s MT tool, the most frequent tagset was verbs in the infinitive compared to its postedited MT output, where verbs in imperative and the second person of plural occurred. These findings are also obtained from the use of the proposed methodology for MT evaluation. The contribution of the proposed methodology is an identification of systematic not random errors. Additionally, the study can also serve as information for optimizing the translation process using postediting.

Detecting packer programs is a key step in the defense against malicious programs and plays a key role in cyber security defenses. One challenge with packer classification is that many features have been used and their individual significance is unknown. An effective approach for building classifiers without requiring prior knowledge of feature significance is to use associative classification (AC) algorithms, which combine association rules and classification. This work considers many different AC algorithms for the challenge of packer detection. Novel variations of many of these algorithms are also developed to address challenges related to having many features of varying types. The effectiveness of the classifiers produced by these algorithms is evaluated, including over time as packers and malware evolve.

The paper focuses on investigating the impact of artificial agent (machine translator) on human agent (posteditor) using a proposed methodology, which is based on language complexity measures, POS tags, frequent tagsets, association rules, and their summarization. We examine this impact from the point of view of language complexity in terms of word and sentence structure. By the proposed methodology, we analyzed 24 733 tags of English to Slovak translations of technical texts, corresponding to the output of two MT systems (Google Translate and the European Commission’s MT tool). We used both manual (adequacy and fluency) and semiautomatic (HTER metric) MT evaluation measures as the criteria for validity. We show that the proposed methodology is valid based on the evaluation of frequent tagsets and rules of MT outputs produced by Google Translate or of the European Commission’s MT tool, and both postedited MT (PEMT) outputs using baseline methods. Our results have also shown that PEMT output produced by Google Translate is characterized by more frequent tagsets such as verbs in the infinitive with modal verbs compared to its MT output, which is characterized by masculine, inanimate nouns in locative of singular. In the MT output, produced by the European Commission’s MT tool, the most frequent tagset was verbs in the infinitive compared to its postedited MT output, where verbs in imperative and the second person of plural occurred. These findings are also obtained from the use of the proposed methodology for MT evaluation. The contribution of the proposed methodology is an identification of systematic not random errors. Additionally, the study can also serve as information for optimizing the translation process using postediting.

The paper focuses on investigating the impact of artificial agent (machine translator) on human agent (posteditor) using a proposed methodology, which is based on language complexity measures, POS tags, frequent tagsets, association rules, and their summarization. We examine this impact from the point of view of language complexity in terms of word and sentence structure. By the proposed methodology, we analyzed 24 733 tags of English to Slovak translations of technical texts, corresponding to the output of two MT systems (Google Translate and the European Commission’s MT tool). We used both manual (adequacy and fluency) and semiautomatic (HTER metric) MT evaluation measures as the criteria for validity. We show that the proposed methodology is valid based on the evaluation of frequent tagsets and rules of MT outputs produced by Google Translate or of the European Commission’s MT tool, and both postedited MT (PEMT) outputs using baseline methods. Our results have also shown that PEMT output produced by Google Translate is characterized by more frequent tagsets such as verbs in the infinitive with modal verbs compared to its MT output, which is characterized by masculine, inanimate nouns in locative of singular. In the MT output, produced by the European Commission’s MT tool, the most frequent tagset was verbs in the infinitive compared to its postedited MT output, where verbs in imperative and the second person of plural occurred. These findings are also obtained from the use of the proposed methodology for MT evaluation. The contribution of the proposed methodology is an identification of systematic not random errors. Additionally, the study can also serve as information for optimizing the translation process using postediting.

We propose a novel scalable approach for testing non-testable programs denoted as ARMED testing. The approach leverages efficient Association Rules Mining algorithms to determine relevant implication relations among features and actions observed while the system is in operation. These relations are used as the specification of positive and negative tests, allowing for identifying plausible or suspicious behaviors: for those cases when oracles are inherently unknownable, such as in social testing, ARMED testing introduces the novel concept of testing for plausibility. To illustrate the approach we walk-through an application example.

The global wine-making industry has faced challenges due to the increasing demands of consumers, particularly in emerging markets such as China, Brazil, India, and Russia. Controlling the quality during wine production is one of the key challenges faced by global winemakers to produce wine with appropriate sensorial properties tailored to specific markets. The wine production quality is constituted from a number of environmental factors such as climate, soil, and temperature, which affect the sensorial properties and the overall quality. This paper proposed a rule-based quality analytics system (RBQAS) to capture physicochemical data during wine production and to investigate the hidden patterns from the data for quality prediction. It consists of IoT for data capture on a real-time basis, followed by association rule mining to identify relationships between sensorial and physicochemical properties of wine.

Mobile databases are becoming more available and thus are drawing more attention from both research and industrial communities. They are currently being widely used in devices such as cell phones, hand-held devices, and notebook computers, among others. Broadcasting is a scalable way to send data from a server to multiple clients. Broadcasting algorithms must be constructed in a way that minimizes the average waiting time for clients. XML is a new standard for representing data in a hierarchical structure, and has many advantages over relational representations due to its portability, flexibility, readability, and customizability. XML has recently been deployed onto many mobile devices; thus a new kind of broadcasting algorithm should be constructed to address the unique characteristics of the way XML databases are queried and accessed. In this paper, we presented a new kind of broadcasting algorithm (BA) by utilizing association-rules in clients' request trends. We implemented three BAs: namely, Exhaustive, Recursive, and Greedy. We tested and compared our BAs with the conventional BAs: namely Sequential and Popularity. The experimental results show that our BAs utilizing association rules perform better than the conventional BAs in both skewed-request situations and requests with association-rules., Dokuz Eylul Univ Presidents Off, TUBITAK, Gonen Bilgi Teknolojileri Danismanlik Ltd

Mobile databases are becoming more available and thus are drawing more attention from both research and industrial communities. They are currently being widely used in devices such as cell phones, hand-held devices, and notebook computers, among others. Broadcasting is a scalable way to send data from a server to multiple clients. Broadcasting algorithms must be constructed in a way that minimizes the average waiting time for clients. XML is a new standard for representing data in a hierarchical structure, and has many advantages over relational representations due to its portability, flexibility, readability, and customizability. XML has recently been deployed onto many mobile devices; thus a new kind of broadcasting algorithm should be constructed to address the unique characteristics of the way XML databases are queried and accessed. In this paper, we presented a new kind of broadcasting algorithm (BA) by utilizing association-rules in clients' request trends. We implemented three BAs: namely, Exhaustive, Recursive, and Greedy. We tested and compared our BAs with the conventional BAs: namely Sequential and Popularity. The experimental results show that our BAs utilizing association rules perform better than the conventional BAs in both skewed-request situations and requests with association-rules., Dokuz Eylul Univ Presidents Off, TUBITAK, Gonen Bilgi Teknolojileri Danismanlik Ltd

There is a well-known lexical gap between content expressed in the form of natural language (NL) texts and content stored in an RDF knowledge base (KB). For tasks such as Information Extraction (IE), this gap needs to be bridged from NL to KB, so that facts extracted from text can be represented in RDF and can then be added to an RDF KB. For tasks such as Natural Language Generation, this gap needs to be bridged from KB to NL, so that facts stored in an RDF KB can be verbalized and read by humans. In this paper we propose LexExMachina, a new methodology that induces correspondences between lexical elements and KB elements by mining class-specific association rules. As an example of such an association rule, consider the rule that predicts that if the text about a person contains the token "Greek", then this person has the relation nationality to the entity Greece. Another rule predicts that if the text about a settlement contains the token "Greek", then this settlement has the relation country to the entity Greece. Such a rule can help in question answering, as it maps an adjective to the relevant KB terms, and it can help in information extraction from text. We propose and empirically investigate a set of 20 types of class-specific association rules together with different interestingness measures to rank them. We apply our method on a loosely-parallel text-data corpus that consists of data from DBpedia and texts from Wikipedia, and evaluate and provide empirical evidence for the utility of the rules for Question Answering.

The relationship between atmospheric components and meteorological variables is essential to assess the air quality and thus avoid citizens' health risks. However, finding an association relationship between those factors could be complicated due to the number of categorization methods that can be used (e.g., frequency, size, binning). Therefore, the objective of this study is to propose a methodology that prepares data through a discretization process and then applies association techniques of the possible combinations between the analyzed variables. The results show that the method used is effective in locating patterns, which are useful for the environmental manager to find knowledge

Knowledge Discovery and Data Mining (KDD) is a multidisciplinary field of study that focuses on methodologies for extracting useful knowledge from data. During the latest Covid-19 pandemic, there was a significant uptick in online-based learning (e-learning) operations as every educational institution moved its operations to digital channels. To increase the quality of education in this new normal, it is necessary to determine the key factors in students’ performance. The main objective of this study is to exploit the regulating factors of education via digital platforms during the covid-19 pandemic by extracting knowledge and a set of rules by using the Decision Tree (j48) classifier.  In this study, we developed a conceptual framework using four datasets, each with a different set of attributes and instances, collected from “X-University” and Microsoft teams. ‘Final term’ and ‘Mid-term’ examinations acted as the root node for all four datasets. The findings of this study would benefit higher education institutions by helping instructors and students to recognize the shortcomings and influences controlling students' performance in the online platforms during the covid-19 pandemic, as well as serve as an early warning framework for predicting students' deficiencies and low school performance.

Given the huge amount of heterogeneous data stored in different locations, it needs to be federated and semantically interconnected for further use. This paper introduces WINFRA, a comprehensive open-access platform for semantic web data and advanced analytics based on natural language processing (NLP) and data mining techniques (e.g., association rules, clustering, classification based on associations). The system is designed to facilitate federated data analysis, knowledge discovery, information retrieval, and new techniques to deal with semantic web and knowledge graph representation. The processing step integrates data from multiple sources virtually by creating virtual databases. Afterwards, the developed RDF Generator is built to generate RDF files for different data sources, together with SPARQL queries, to support semantic data search and knowledge graph representation. Furthermore, some application cases are provided to demonstrate how it facilitates advanced data analytics over semantic data and showcase our proposed approach toward semantic association rules.

Higher Education institutions mainly aim at improving the teaching-learning process. In this context, the in-terrelationship among courses in the development of the Curricular Design becomes essential. However, in many cases not all the required efforts are made in order for subjects to be effectively integrated. This paper presents a learning experience designed for a core subject, Artificial Intelligence, with the purpose of allowing students to integrate knowledge through practical activities as well as achieving better articu-lation with previously acquired academic contents throughout their course of study., Mejorar el proceso de enseñanza-aprendizaje es uno de los objetivos que tienen las instituciones de educa-ción superior. En este contexto, la interrelación entre las asignaturas del diseño curricular se torna impres-cindible. Sin embargo, muchas veces no se realizan los esfuerzos necesarios para integrarlas efectivamen-te. En este trabajo se presenta una experiencia para una asignatura troncal, Inteligencia Artificial, la cual permite al alumno integrar conocimientos median-te actividades prácticas y lograr así la articulación con algunos contenidos académicos adquiridos en el transcurso de la carrera.

Association rules are rules that define relationships between items in sales databases. They have been used primarily to organize relevant products in stores in a way to makes them more visible to consumers, which may increase sales and profits. On the other hand, it has been rarely used in recommender systems where algorithms provide instant recommendations by processing consumers’ interests that are gathered when browsing online. However, the vast amount of information collected from transaction data saved on backup servers is poorly taken advantage of, because it is not connected to the Internet, although interesting and personalized recommendations can be created after finding the collections of most frequent items, or most interesting rules in such databases. In this paper, we do a critique of the existing research on both recommender systems along with showing their drawbacks, and the association rules with detailed explanations on their advantages. Finally, draw up with several solutions for producing high quality as well as accurate recommendations by applying novel combinations of techniques observed in this research area including the association-rules-based recommender systems.

In this fast information-technological world, data grow bigger and bigger. Big data refer to the huge volume, high velocity, wide variety and different veracity of the data flow. One of the big challenges is to aggregate and visualize the data mining results to the end user. To achieve a certain level of data validity for better data aggregation in estimation and prediction, the synergy of data visualization models and data mining strategies are necessary. In this MSc thesis, the focus is on frequent pattern visualization. This aims to visualize textual frequent patterns, which result in graphical representation with three important information: frequency distribution, cardinality distribution and superset-subset relationship. The designed and implemented visualization system for frequent pattern mining reveals interesting information and useful knowledge mined from the transactional databases for various applications and services.

The purpose of this study is to analyse the suitability of photo-sharing platforms, such as Flickr, to extract relevant knowledge on tourists’ spatial movement and point of interest (POI) visitation behaviour and compare the most prominent clustering approaches to identify POIs in various application scenarios.

Prediction of students' success is one of the topics of research of the area known as the discovery of knowledge from educational data sets. The aim of this dissertation is focused on determining the students' success factors in a blended learning environment by combined implementation of methods for knowledge discovery and in-depth analysis of data. A case study study conducted at the School of Electrical and Computer Engineering of Applied Studies in Belgrade has been described. Available data about the pedagogical process and the academic results of the Computer graphics course enabled comparative analyzes of the realized experiments. The outcome of the research is to establish a methodology for discovering knowledge from a blended learning environment. The conceptual model of prediction is based on realized student activities within the Moodle course and the classical way of teaching.

Today's ultra-connected world is generating a huge amount of data stored in databases and cloud environment especially in the era of transportation. These databases need to be processed and analyzed to extract useful information and present it as a valid element for transportation managers for further use, such as road safety, shipping delays, and shipping optimization. The potential of data mining algorithms is largely untapped, this paper shows large-scale techniques such as associations rule analysis, multiple criteria analysis, and time series to improve road safety by identifying hot-spots in advance and giving chance to drivers to avoid the dangers. Indeed, we proposed a framework DM-MCDA based on association rules mining as a preliminary task to extract relationships between variables related to a road accident, and then integrate multiple criteria analysis to help decision-makers to make their choice of the most relevant rules. The developed system is flexible and allows intuitive creation and execution of different algorithms for an extensive range of road traffic topics. DM-MCDA can be expanded with new topics on demand, rendering knowledge extraction more robust and provide meaningful information that could help in developing suitable policies for decision-makers.

FP-growth algorithm is a classic algorithm of mining frequent item sets, but there exist certain disadvantages for mining the weighted frequent item sets. Based on the weighted downward closure property of the weighted model, this paper proposed a method to reduce the use of storage space by constructing a weight ordered FP-tree, so as to improve the generation efficiency of weighted frequent item sets.

The most common type of cancer in women worldwide is the Breast Cancer. Breast cancer may be detected early using Mammograms, probably before it's spread. Recurrent breast cancer could occur months or years after initial treatment. The cancer could return within the same place because the original cancer (local recurrence), or it may spread to different areas of your body (distant recurrence). Early stage treatment is done not only to cure breast cancer however additionally facilitate in preventing its repetition/recurrence. Data mining algorithms provide assistance in predicting the early-stage breast cancer that continually has been difficult analysis drawback. The projected analysis can establish the most effective algorithm that predicts the recurrence of the breast cancer and improve the accuracy the algorithms. Large information like Clump, Classification, Association Rules, Prediction and Neural Networks, Decision Trees can be analyzed using data mining applications and techniques.

El propósito de este proyecto es investigar en la mejora de las técnicas de extracción de Reglas de Asociación (RA) entre enfermedades, o entre enfermedades y otros conceptos médicos. Estas reglas permiten representar el conocimiento médico subyacente a un conjunto de Historias Clínica Electrónica (HCE). Concretamente nos planteamos explorar técnicas semisupervisadas que nos permitan alcanzar resultados equiparables a los de las técnicas supervisadas con una mínima supervisión. El proyecto se propone realizar avances significativos en la selección de reglas de asociación relevantes en el dominio de la salud, que pueden tener una alta aplicabilidad en la ayuda al diagnóstico y en la prevención de enfermedades., This project aims to improve the techniques for extracting Association Rules (AR) between diseases, or between diseases and other medical concepts. These rules allow the representation of medical knowledge underlying a set of Electronic Medical Records (EHR). Particularly, we plan to explore semi-supervised techniques that allow us to achieve similar results to those obtained using supervised techniques, but requiring minimal supervision. The project intends to make significant progress in the selection of relevant AR, which may be applied in the health domain for developing diagnostic help systems, or for disease prevention.

Despite protected by firewalls and network security systems, databases are vulnerable to attacks especially when the perpetrators are from within the organization and have authorized access to these systems. Detecting their malicious activities is difficult as each database has its own set of unique usage activities and the generic exploitation avoidance rules are usually not applicable. This paper proposes a novel method to improve the security of a database by using machine learning to learn the user behaviour unique to a database environment and apply that learning to detect anomalous user activities through the analysis of sequences of user session data. Once these suspicious users are detected, their privileges are systematically suppressed. The empirical analysis shows that the proposed approach can intuitively adapt to any database that supports a wide variety of clients and enforce stringent control customized to the specific ITsystems.

The widespread use of the Internet and the frequency with which people access websites, particularly those ones dedicated to electronic commerce, have given rise to a need to know how users interact with them. This situation leads to a huge amount of information being generated daily that can become a key factor for companies as long as it is analyzed properly. That is why, this work focuses on the analysis of search patterns performed on e-commerce websites based on Web Analytics and Data Science. With this aim, a data set has been generated to which the selected algorithm has been applied to find association rules. In particular, the Google Analytics tool has been applied in the study. As a result, patterns of searches that users of an e-commerce website could perform were found., La masificación del uso del Internet y la frecuencia con la que las personas acceden a sitios web, particularmente los dedicados al comercio electrónico, han dado origen a una necesidad de conocer la forma en que los usuarios interactúan con ellos. Esta situación conduce a que diariamente se genere una cantidad enorme de información que puede convertirse en un factor clave para las empresas siempre y cuando sea analizada adecuadamente. Es por ello que, este trabajo se enfoca al análisis de patrones de búsquedas realizadas en sitios web de comercio electrónico basado en Web Analytics y Data Science. Para ello, se ha generado un conjunto de datos al que se le ha aplicado el algoritmo seleccionado para encontrar reglas de asociación. Particularmente, se ha aplicado la herramienta Google Analytics en el estudio. Como resultado se han encontrado patrones de búsquedas que podrían realizar los usuarios de un sitio web de comercio electrónico.

In the present paper we aim to study the visual decision support based on Cellular machine CASI (Cellular Automata for Symbolic Induction). The purpose is to improve the visualization of large sets of association rules, in order to perform Clinical decision support system and decrease doctors’ cognitive charge. One of the major problems in processing association rules is the exponential growth of generated rules volume which impacts doctor’s adaptation. In order to clarify it, many approaches meant to represent this set of association rules under visual context have been suggested. In this article we suggest to use jointly the CASI cellular machine and the colored 2D matrices to improve the visualization of association rules. Our approach has been divided into four important phases: (1) Data preparation, (2) Extracting association rules, (3) Boolean modeling of the rules base (4) 2D visualization colored by Boolean inferences.

En el presente escrito se aborda el estudio de la metodología denominada análisis cesta de la compra, una técnica perteneciente a la minería de datos. Desarrollada en un principio para ser aplicada en el ámbito del marketing, su aplicación ha traspasado a campos de investigación más diversos. La finalidad que se persigue es identificar aso ciaciones entre diversos grupos de productos o categorías. Además de validar asocia ciones de productos que pueden considerarse obvias, debido a su reiterada compra conjunta, el análisis cesta de la compra busca la contingencia entre productos cuya asociación no es evidente. Es asimismo destacable el uso que se realiza de datos considerados como inservibles y desordenados en la gestión de búsqueda de asociaciones; siendo la principal fuente de información disponible la base de datos que recoge las transacciones que realizan los consumidores en cada compra., The current essay offers an approach to the market basket analysis metodology, known to be a technique to mining data. Once conceived as a technique with an exclusive use on marketing studies, its applications have knowadays moved on to a broader spectrum of investigation fields such as biotechnology. Its main purpose is to identify associations between different groups or categories of products. Asside from obvious associations be tween those products that are often purchased concurrently, this analysis seeks the contingency across products that have a not-so-defined association at first. It is key to highlight the way this analytical method exploids aparrently unuseful data, building from raw data a structured set of analytcal information about customer's trans actions on each of their purchases.

The behaviour of a library user is based on his ability to evaluate the consumption of library services based on alternative market services. The purpose of this paper is to study the usage behaviour in a public library building in the context of alternative costs instead of a narrow focus on book circulation data. Design/methodology/approach - By library usage mining, using associative rules, the authors described the behaviour of library users and identified the typical behaviour during the visits. The authors analyse the results in the context of alternative costs assigned to the visits. Findings - The results confirm that some underused services, such as digital services, deliver significantly greater benefit. The frequency of use, the duration of visit and the number of items used are associated with higher alternative costs. There were no significant differences in alternative cost within economic groups (excluding pensioners). This paper identified 41 frequent patterns with different alternative costs., Chování uživatelů knihovních služeb je založeno na jejich schopnosti hodnotit užitek ze spotřeby. Toto hodnocení je založeno na schopnosti analyzovat i tzv. alternativní nákladů k zajištění substitutů. Cílem příspěvku je proto analyzovat a studovat chování čtenářů v knihovně a to v kontextu alternativních nákladů. Jde o zcela nový přístup, který rozšiřuje možnosti analýzy chování spotřebitelů kulturních služeb.

En la era digital en la que nos encontramos, las empresas cada vez más cuentan con grandes cantidades de datos de sus clientes. Esto ha sido posible gracias al avance exponencial de la tecnología y al abaratamiento del almacenamiento de datos. Todo ello ha provocado la necesidad en las empresas de aprender a tratar estos datos y poder obtener de ellos información útil que les permita conocer mejor aún a sus clientes. Existen multitud de técnicas de análisis de datos que permiten obtener la información oculta en las grandes bases de datos, información que puede proporcionar a la empresa una gran ventaja competitiva en el mercado. En este estudio vamos a profundizar en una de esas técnicas, las reglas de asociación, mostrando al lector el potencial que tiene una técnica tan sencilla como esta, In the current digital era, companies increasingly have amounts of data from their customers. This has been possible thanks to the exponential advance of technology and the cheapening of data storage. For this reason, the companies need to learn about data manipulation and to obtain useful information that allows them get to know their customers better. There are multitude of data analysis techniques that allow to obtain information hidden in large databases, information that can provide the company a great competitive advantage in the market. In this study, we will delve into one of these techniques, the association rules, showing the reader the potential of a simple technique like this.

В результаті роботи було розроблено систему моніторингу додатків з мікросервісною архітектурою. Основними перевагами системи є наявність засобів інтеграції в процес неперервної доставки програмного забезпечення та підтримка додатків з мікросервісною архітектурою., The result of the work is a performance monitoring system for applications with microservice architecture. The main advantages of the system include means for simple integration into the process of continuous software delivery and support of applications with microservice architectures., Результатом работы является система мониторинга приложений с микросервисной архитектурой. Основными преимуществами системы является наличие средств интеграции в процесс непрерывной доставки программного обеспечения, а также поддержка приложений с микросервисной архитектурой.

[EN] "This article might interest you", "People who bought this article, also bought ..." or "This might also interest you" are phrases that are increasingly present in the daily activity of users that consume products and services in virtual stores and other services. The reason for it is that recommendation systems have been consolidated as a strong trend for the growth of digital commerce in recent years, especially with regards to big data. Advances and the cheapening of technologies have allowed many companies to build their virtual environments to complement their physical store, or they were even created in a purely virtual plan and, as a result, in both cases, the interactions of customers with their virtual environments have stored a huge amount of data about their preferences, such as searched and / or purchased products, movies or songs played or marked as favorites, or even events captured by sensors and registered in a database, since the Internet Things is an agent present and in great ascension that records data about people and their choices involuntarily. The main objective of this work is to use the existing data about users’ preferences, and apply Machine Learning techniques to develop a system to make recommendations, suggesting new items adjusted to users’ tastes based on users’ profiles or products generated for this purpose. Currently, several service providers already offer SaaS solutions (Software as a Service) to companies in order to integrate personalized recommendations into their commercial project. Among the best known are BrainSINS7 , Barilliance8 or Certona9 . Unlike the service offered by these providers, our proposal is very non-invasive in the sense that it only uses the data of the clients' election history, without other data about them being necessary. To do this, we propose using a hybrid method that combines collaborative and item-to-item filtering. Besides security due to customers’ anonymity and for not demanding the challe, [ES] “Este artículo podría gustarle”, “Personas que compraron este artículo también compraron …” o “Esto también podría interesarle” son frases que están cada vez más presentes en la actividad cotidiana de los usuarios consumidores de productos y servicios en tiendas virtuales y otros servicios. Esto se debe a que los sistemas de recomendación se han consolidado como fuerte tendencia para el crecimiento del comercio digital en los últimos años, sobre todo en relación con el big data. Los avances y el abaratamiento de las tecnologías permitieron que muchas empresas construyesen sus entornos virtuales para complementar a su tienda física, o que fuesen creadas en plan puramente virtual y, resulta que, en ambos casos, las interacciones de los clientes con sus entornos virtuales han generado una cantidad enorme de datos acerca de sus preferencias, como por ejemplo, productos buscados y/o comprados, películas o canciones reproducidas o marcadas como favoritas, o incluso eventos capturados por sensores y registrados en una base de datos, dado que la Internet de las Cosas es un agente presente y en plena ascensión que registra datos sobre las personas y sus elecciones de forma involuntaria. Este trabajo tiene como objetivo principal usar los datos ya existentes sobre las preferencias de los usuarios, y aplicar técnicas de Machine Learning para desarrollar un sistema para hacer recomendaciones, sugiriendo nuevos ítems ajustados a los gustos de los usuarios a partir de perfiles de usuario o de productos generados con este fin. Actualmente, diversos proveedores ya ofrecen a las empresas soluciones SaaS (Software as a Service) con las que integrar recomendaciones personalizadas en su proyecto comercial. Entre los más conocidos se encuentran BrainSINS1 , Barilliance2 o Certona3 . A diferencia del servicio ofrecido por estos proveedores, nuestra propuesta es muy poco invasiva en el sentido que usa únicamente los datos del historial de elecciones de los clientes, sin que sean nece, [CA] “Aquest article podria agradar-li”, “Persones que van comprar aquest article també van comprar …” o “Açò també podria interessar-li” són frases que estan cada vegada més presents en l'activitat quotidiana dels usuaris consumidors de productes i serveis en tendes virtuals i altres serveis. Açò es deu al fet que els sistemes de recomanació s'han consolidat com a forta tendència per al creixement del comerç digital en els últims anys, sobretot en relació amb el big data. Els avanços i l'abaratiment de les tecnologies van permetre que moltes empreses construïren els seus entorns virtuals per a complementar a la seua tenda física, o que anaren creades en pla purament virtual i, resulta que, en tots dos casos, les interaccions dels clients amb els seus entorns virtuals han generat una quantitat enorme de dades sobre les seues preferències, com per exemple, productes cercats i/o comprats, pel·lícules o cançons reproduïdes o marcades com a favorites, o fins i tot esdeveniments capturats per sensors i registrats en una base de dades, atès que la Internet de les Coses és un agent present i en plena ascensió que registra dades sobre les persones i les seues eleccions de forma involuntària. Aquest treball té com a objectiu principal usar les dades ja existents sobre les preferències dels usuaris, i aplicar tècniques de Machine Learning per a desenvolupar un sistema per a fer recomanacions, suggerint nous ítems ajustats als gustos dels usuaris a partir de perfils d'usuari o de productes generats a aquest efecte. Actualment, diversos proveïdors ja ofereixen a les empreses soluciones SaaS (Program as a Service) amb les quals integrar recomanacions personalitzades en el seu projecte comercial. Entre els més coneguts es troben BrainSINS4 , Barilliance5 o Certona6 . A diferència del servei oferit per aquests proveïdors, la nostra proposta és molt poc invasiva en el sentit que usa únicament les dades de l'historial d'eleccions dels clients, sense que siguen necessaris altres dade

An analysis of occupational accidents in the mining sector was conducted using the data from the Spanish Ministry of Employment and Social Safety between 2005 and 2015, and data-mining techniques were applied. Data was processed with the softwareWeka. Two scenarios were chosen from the accidents database: surface and underground mining. The most important variables involved in occupational accidents and their association rules were determined. These rules are composed of several predictor variables that cause accidents, defining its characteristics and context. This study exposes the 20 most important association rules in the sector—either surface or underground mining—based on the statistical confidence levels of each rule as obtained byWeka. The outcomes display the most typical immediate causes, along with the percentage of accidents with a basis in each association rule. The most important immediate cause is body movement with physical effort or overexertion, and the type of accident is physical effort or overexertion. On the other hand, the second most important immediate cause and type of accident are different between the two scenarios. Data-mining techniques were chosen as a useful tool to find out the root cause of the accidents., Peer Reviewed, Postprint (published version)

This technical report presents example cases for the algorithm introduced in our paper Risk Mediation in Association Rules: The Case of Decision Support in Medication Review.

We propose a model for the incorporation of risk in association rule application. We validate this model using data gathered in a randomized controlled trial from a recommender system for medication reviews in primary care. The model’s outcomes are found to have predictive value when tested against decisions made by physicians on 261 patients’ health records.

Osnovni cilj istraživanja Doktorske disertacije je definisanje okvira za sprovođenje celovitog istraživačkog poduhvata unapređenja modela poslovnog odlučivanja i otkrivanja zakonitosti u podacima za potrebe brojnih analiza: pre svega otkrivanja asocijativnih pravila i predviđanja, kao i upotrebe rezultata radi donošenja ispravnih upravljačkih poslovnih odluka. Dakle, cilj je analiza i primena sistema asocijativnih pravila radi unapređenja modela poslovnog odlučivanja menadžera najvišeg nivoa poslovnog sistema, radi donošenja efektivnih i efikasnih odluka. U istraživanju su korišćene savremene naučne metode iz oblasti poslovne inteligencije. Glavna hipoteza: “Moguće je unaprediti model poslovnog odlučivanja sistemom asocijativnih pravila“ je potvrđena u istraživanju. Ukazano je na značaj poslovne inteligencije za stvaranje modela koji može povećati efektivnost procesa menadžerskog odlučivanja. Primena asocijativnih pravila u svrhe istraživanja ima izuzetan potencijal u oblasti poslovanja. Razvijen je i prikazan model poslovnog odlučivanja pomoću sistema asocijativnih pravila. Dokazano je da je ova oblast poslovne inteligencije veoma aktuelna i sa velikim potencijalom. Izvedeni zu zaključci i date su smernice za buduća istraživanja kao izazov da se pruže značajan naučni i stručni doprinos sa ciljem unapređenja poslovnog odlučivanja., Main goal of research of the Doctoral Dissertation is defining the framework for integral research project of improvement of model of decisionmaking and data mining for numerous analysis: mining of association rules and prediction, as well as the use of results in order to gain effective management decisions. The goal of the research is analysis and application of system of association rules in order to improve the model of business decision-making of top-level managers of business system, in order to get the most effective decisions. Modern scientific methods from the field of business intelligence have been used during the research. The main hypothesis: “It is possible to improve the model of business decicion-making by system of association rules” has been confirmed during the research. The importance of business intelligence for creation of model that can increase the effectiveness of managers’ decision making is highlighted. The application of association rules with the purposes of research has an immense potential in the business. The model of business decision-making by association rules has been developed and presented. It is proven that this field of business intelligence is very popular and has big potential. Concluding remarks, as well as the recommendations for future research have been given in order to provide significant scientific and professional contribution with goal of improving business decision- making.

Sensors are widely used in all aspects of our daily life including factories, hospitals and even our homes. Discovering time series association rules from sensor data can reveal the potential relationship between different sensors which can be used in many applications. However, the time series association rule mining algorithms usually produce rules much more than expected. It's hardly to understand, present or make use of the rules. So we need to prune and summarize the huge amount of rules. In this paper, a two-step pruning method is proposed to reduce both the number and redundancy in the large set of time series rules. Besides, we put forward the BIGBAR summarizing method to summarize the rules and present the results intuitively., DAWN, SMART (Smarta system och tjänster för ett effektivt och innovativt samhälle)