Categorical Clustering Research Articles

Digital Divide in the European Union: A Typology of EU Citizens

AbstractThis paper addresses the heterogeneity of the digital divide and internet use among citizens in the 28 European Union (EU) countries (at the time of the survey). Drawing from the Eurobarometer Surveys, three indicators of the digital divide are used to define the groups: frequency of internet access, means of internet access, and online activities. The categorical clustering algorithm identifies six groups of internet users: Non-Users, Basic Users, Information Exchangers, Instrumental Users, Socializers, and Advanced Users, each with distinct socio-demographic profiles. The study reveals significant socio-economic and demographic profiling variables characterizing these patterns, including age, education, gender, occupation, type of community and geographic location. A major digital divide is detected in many countries; Notably, Romania, Greece, and Bulgaria have the largest proportion of Non-Users, emphasizing the need for targeted policy interventions. These results provide crucial insights for the European Commission’s digitization strategy, suggesting that more nuanced and targeted measures are needed to ensure equitable digital access across the EU.

Interpretable sequence clustering

Categorical sequence clustering is vital across various domains; however, the interpretability of cluster assignments presents considerable challenges. Sequences inherently lack explicit features, and existing sequence clustering algorithms heavily rely on complex representations, which complicates the explanation of their outcomes. To address this issue, we propose a method called Interpretable Sequence Clustering Tree (ISCT), which combines sequential patterns with a concise and interpretable tree structure. ISCT leverages k−1 patterns to generate k leaf nodes, corresponding to k clusters, which provides an intuitive explanation on how each cluster is formed. More precisely, ISCT first projects sequences into random subspaces and then utilizes the k-means algorithm to obtain high-quality initial cluster assignments. Subsequently, it constructs a pattern-based decision tree using a boosting strategy in which sequences are re-projected and re-clustered at each node before mining the top-1 discriminative splitting pattern. Experimental results on 14 real-world data sets demonstrate that our proposed method provides an interpretable tree structure while delivering fast and accurate cluster assignments.

What and how? Or why and for whom? A content analysis of middle school general music literature

The purpose of this content analysis is to analyze available peer-reviewed middle school general music literature to determine the ways in which middle school general music has been conceived of in extant literature in the United States. The researchers performed a content analysis of peer-reviewed empirical and practitioner articles relating to middle school general music ( N = 45), creating a coding scheme of categorical clusters that related to the Association of Middle Level Education’s position paper: This We Believe. Content coding revealed curricular content (78%) and pedagogical practice (42%) as the most common topics. Several topics were underrepresented in the literature including student opinion/voice (11%), inclusive practice (18%), adolescent development (20%), cultural awareness (18%), classroom culture (20%), and teacher preparation and development (16%). These findings may have important implications for preservice music teacher educators preparing teachers for the context of middle school general music, as well as for researchers within this domain.

Unsupervised recognition of radar signals combining multi-block TFR with subspace clustering

In the realm of radar systems, the proliferation of new modulation techniques introduces an increased level of complexity in the identification of both existing and potentially novel modulation schemes. Conventional supervised recognition methodologies, which rely heavily on labeled datasets, exhibit limitations in distinguishing amongst a multitude of unlabeled signals. This paper introduces a novel framework that synergizes Subspace Clustering (SC) with Multi-Block Time-Frequency Representations (TFRs, denoted as mT), referred to as SC-mT. This framework leverages subspace clustering for the unsupervised categorization of radar modulations and employs an innovative multi-block strategy to augment classification precision. The process commences with the generation of TFR datasets for radar signals, followed by the construction of two distinct multi-block models, Model-A and Model-B. These models segment the radar TFRs into overlapping multi-block sets, utilizing the concept of random receptive fields. The subspace clustering algorithm is then applied to each block within the sub-TFR sets to procure an affinity matrix. The culmination of this process involves the aggregation of the affinity matrices from all blocks, facilitating the derivation of classification outcomes via spectral clustering. Empirical analyses affirm that the sequential integration of the SC-mT algorithm surpasses the classification efficacy of various conventional and cutting-edge algorithms. Notably, this algorithm attains a classification accuracy surpassing 90% for ten unlabeled signals, even in scenarios where the signal-to-noise ratio (SNR) is as low as -4 dB.

Categorical Data Clustering with K-Modes Method on Fire Cases in DKI Jakarta Province

In DKI Jakarta Province, the number of fires increases and decreases every year. For this reason, efforts need to be made to prevent and reduce the risk of fire. BPBD DKI Jakarta is responsible for this matter. However, for these efforts to be effective, information is needed regarding fire patterns that frequently occur. Fire patterns can be seen using K-Modes categorical clustering analysis. The data used is fire data in DKI Jakarta in 2018. The optimal number of clusters was obtained as 6 clusters based on the Davies Bouldin Index value with the smallest DBI value is 6,22. Of the six clusters, cluster 3 is the cluster with the highest number of fire cases. Cluster 3 has a centroid, namely that fire cases occurred on Friday, November, in Cakung District, due to an electrical short circuit, burning down residential houses and rarely causing minor injuries, serious injuries or deaths.

Cultivation of human centered artificial intelligence: culturally adaptive thinking in education (CATE) for AI.

Artificial Intelligence (AI) has become ubiquitous in human society, and yet vast segments of the global population have no, little, or counterproductive information about AI. It is necessary to teach AI topics on a mass scale. While there is a rush to implement academic initiatives, scant attention has been paid to the unique challenges of teaching AI curricula to a global and culturally diverse audience with varying expectations of privacy, technological autonomy, risk preference, and knowledge sharing. Our study fills this void by focusing on AI elements in a new framework titled Culturally Adaptive Thinking in Education for AI (CATE-AI) to enable teaching AI concepts to culturally diverse learners. Failure to contextualize and sensitize AI education to culture and other categorical human-thought clusters, can lead to several undesirable effects including confusion, AI-phobia, cultural biases to AI, increased resistance toward AI technologies and AI education. We discuss and integrate human behavior theories, AI applications research, educational frameworks, and human centered AI principles to articulate CATE-AI. In the first part of this paper, we present the development a significantly enhanced version of CATE. In the second part, we explore textual data from AI related news articles to generate insights that lay the foundation for CATE-AI, and support our findings. The CATE-AI framework can help learners study artificial intelligence topics more effectively by serving as a basis for adapting and contextualizing AI to their sociocultural needs.

Reproducible clustering with non-Euclidean distances: a simulation and case study

AbstractCertain categorical sequence clustering applications require path connectivity, such as the clustering of DNA, click-paths through web-user sessions, or paths of care clustering with sequences of patient medical billing codes. $$K$$ K -means and $$k$$ k -medoids clustering with non-Euclidean distance metrics such as the Jaccard or edit distances maintains such path connectivity. Although $$k$$ k -means and $$k$$ k -medoids clustering with the Jaccard and edit distances have enjoyed success in these domains, the limits of accurate cluster recovery in these conditions have not yet been defined. As a first step in approaching this goal, we performed a simulated study using $$k$$ k -means and $$k$$ k -medoids clustering with non-Euclidean distances and show the performance deteriorates at a certain level of noise and when the number of clusters increases. However, we identify initialization strategies that improve upon cluster recovery in the presence of noise. We employ the use of the Tibshirani and Guenther (J Comput Graph Stat 14(3):511–528, 2005) Prediction Strength method, which creates a hypothesis testing scenario that determines if there is clustering structure to the data (if the clusters are reproducible), with the null hypothesis being there is none. We then applied the framework to perinatal episodes of care and the clusters reproducibly and organically split between Cesarean and vaginal deliveries, which itself is not a clinical finding but sensibly validates the approach. Further visualizations of the clusters did bring insights into subclusters that split along groups of physicians, cost and risk scores, warranting the outlined future work into ways of improving this framework for better resolution.

Parameterized complexity of categorical clustering with size constraints

In the Categorical Clustering problem, we are given a set of vectors (matrix) A={a1,…,an} over Σm, where Σ is a finite alphabet, and integers k and B. The task is to partition A into k clusters such that the median objective of the clustering in the Hamming norm is at most B. Fomin, Golovach, and Panolan [ICALP 2018] proved that the problem is fixed-parameter tractable for the binary case Σ={0,1}. We extend this algorithmic result to a popular capacitated clustering model, where in addition the sizes of the clusters are lower and upper bounded by integer parameters p and q, respectively. Our main theorem is that the problem is solvable in time 2O(Blog⁡B)|Σ|B⋅(mn)O(1).

기억술을 활용한 교양 음악 수업 방안

Objectives The purposes of this study is to explore how mnemonic methods can be applied to liberal arts music classes by exploring the various mnemonic methods and their applicability. Methods To this end, based on past studies on mnemonics, possible ways to apply such methods to composition activities in introductory music classes were explored by identifying different categories and methods that could be adapted. Results Composition activities using mnemonics were into two steps: preparing the score (remembering the order of preparing the score, writing clefs, writing key signatures, and writing time signatures) and the basic composition process (memorizing composition rules based on the 3 major chords, memorizing notes and beats, memorizing note names, memorizing tempo, and memorizing ellipsis). To this end, a total of eight mnemonic methods (story method, acrostics, categorical clustering, key word method, gesture, acronyms, rhyme, and peg word method) were used. Conclusions This study can be meaningful in that by making it easy to memorize the contents that are difficult to remember in the basic music learning process through the use of mnemonics, It can be additionally meaningful in that it can increase the use of various mnemonic methods to help in sheet music reading skills which can allow composition activities.

SegMix: Co-occurrence Driven Mixup for Semantic Segmentation and Adversarial Robustness

In this paper, we present a strategy for training convolutional neural networks to effectively resolve interference arising from competing hypotheses relating to inter-categorical information throughout the network. In this work, this is accomplished for the task of dense image labelling by blending images based on (i) categorical clustering or (ii) the co-occurrence likelihood of categories. We then train a source separation network which simultaneously segments and separates the blended images. Subsequent feature denoising to suppress noisy activations reveals additional desirable properties and high degrees of successful predictions. Through this process, we reveal a general mechanism, distinct from any prior methods, for boosting the performance of the base segmentation and salient object detection network while simultaneously increasing robustness to adversarial attacks.

Toward a More Personalized MOOC: Data Analysis to Identify Drinking Water Production Operators’ Learning Characteristics—An Ecuador Case

Only 35% of the Ecuadorian population consumes drinking water of “assured quality”. One of the causes is related to the deficiencies in the technical ability of the operators due to their lack of education, technical training, and experience. Massive open online courses (MOOCs) responsive to characteristics and learning needs are an option to strengthen the skills of operators. The goal of the present study is therefore to describe a methodology that includes the application of a survey and the use of statistical methods such as categorical principal component analysis (CATPCA) and cluster analysis to identify and assess learning characteristics. The results present the most frequent variables in the personal, academic, emotional, social, and cognitive aspects. They also show the preferences and learning needs of the operators. Finally, it is concluded that this study identifies common learning characteristics, needs, and preferences that are relevant for the creation of a quality personalized instructional design in MOOCs.

Categorical Data Clustering Using Harmony Search Algorithm for Healthcare Datasets

Healthcare analytics provide many benefits in healthcare dashboard systems. Healthcare datasets majorly contains categorical attributes. This paper proposed an optimized clustering for healthcare dataset named harmony search based categorical clustering (HSCC). The existing k-modes clustering algorithm is one of the well-known categorical data-clustering algorithm. Since the k-modes algorithm produces local optimal clusters. Generally, researchers use genetic algorithm (GA) based clustering algorithms to converge locally optimal solutions to global optimal solutions. GA has some deficiencies such as premature convergence with low speed. In this paper, harmony search (HS) optimization algorithm used to optimize clustering results. The result shows the proposed HSCC algorithm produced global optimized solution, unbiased and matured results. HSCC produces 98% accuracy for dental and 71% for lung cancer dataset. While GACC produces 95% and 65% accuracy for dental dataset and lung cancer dataset.

COIN: Correlation Index-Based Similarity Measure for Clustering Categorical Data

In this paper, a correlation index-based clustering algorithm (COIN) is proposed for clustering the categorical data. The proposed algorithm was tested on nine datasets gathered from the University of California at Irvine (UCI) repository. The experiments were made in two ways, one by specifying the number of clusters and another without specifying the number of clusters. The proposed COIN algorithm is compared with five existing categorical clustering algorithms such as Mean Gain Ratio (MGR), Min–Min-Roughness (MMR), COOLCAT, K-ANMI, and G-ANMI. The result analysis clearly reports that COIN outperforms other algorithms. It produced better accuracies for eight datasets (88.89%) and slightly lower accuracy for one dataset (11%) when compared individually with MMR, K-ANMI, and MGR algorithms. It produced better accuracies for all nine datasets (100%) when it is compared with G-ANMI and COOLCAT algorithms. When COIN was executed without specifying the number of clusters, it outperformed MGR for 88.89% of the test instances and produced lower accuracy for 11% of the test instances.

Rough set based information theoretic approach for clustering uncertain categorical data

Many real applications such as businesses and health generate large categorical datasets with uncertainty. A fundamental task is to efficiently discover hidden and non-trivial patterns from such large uncertain categorical datasets. Since the exact value of an attribute is often unknown in uncertain categorical datasets, conventional clustering analysis algorithms do not provide a suitable means for dealing with categorical data, uncertainty, and stability. The ability of decision making in the presence of vagueness and uncertainty in data can be handled using Rough Set Theory. Though, recent categorical clustering techniques based on Rough Set Theory help but they suffer from low accuracy, high computational complexity, and generalizability especially on data sets where they sometimes fail or hardly select their best clustering attribute. The main objective of this research is to propose a new information theoretic based Rough Purity Approach (RPA). Another objective of this work is to handle the problems of traditional Rough Set Theory based categorical clustering techniques. Hence, the ultimate goal is to cluster uncertain categorical datasets efficiently in terms of the performance, generalizability and computational complexity. The RPA takes into consideration information-theoretic attribute purity of the categorical-valued information systems. Several extensive experiments are conducted to evaluate the efficiency of RPA using a real Supplier Base Management (SBM) and six benchmark UCI datasets. The proposed RPA is also compared with several recent categorical data clustering techniques. The experimental results show that RPA outperforms the baseline algorithms. The significant percentage improvement with respect to time (66.70%), iterations (83.13%), purity (10.53%), entropy (14%), and accuracy (12.15%) as well as Rough Accuracy of clusters show that RPA is suitable for practical usage. We conclude that as compared to other techniques, the attribute purity of categorical-valued information systems can better cluster the data. Hence, RPA technique can be recommended for large scale clustering in multiple domains and its performance can be enhanced for further research.

Rough set based on least dissimilarity normalized index for handling uncertainty during E-learners learning pattern recognition

The determination of e-learners' learning style in an online environment has raised the potential scope of interest as its exact estimation prompts a sensational improvement in the contents of the learning framework and student performance. It requires a deep investigation of the learning habits of the learner. Grouping e-learners together provides a more quantifiable way to analyze the learner's feedback and log files to discriminate them based on their learning style. This is accomplished with the help of clustering algorithms in data mining that aids in determining their learning styles well. The target clusters are analyzed by generating functional patterns or rules using the rule induction algorithms. Most of the existing works in the literature attributed to the elucidation of learning styles fail to address the uncertainty and inconsistency in the learner's characteristics. The RST is an optimal method for analyzing the learner's behavior in this context. Thus, a Rough set based least dissimilarity normalized index (RS-LDNI) is proposed for resolving uncertainty while estimating e-learners' learning patterns. This RS-LNDI used the merits of Maximum Dependency Attributes (MDA) for categorical clustering such that the maximal dependency between attributes can be determined by splitting attributes instead of Roughness. It also adopted categorical data clustering to attain the correlation between attributes that cannot be used for learning style prediction. The experimental results of the RS-LNDI algorithm outperform the demerits of these existing clustering algorithms by utilizing the reduct and equivalence class property of rough set theory.

Semi-Supervised Domain Adaptive Structure Learning.

Semi-supervised domain adaptation (SSDA) is quite a challenging problem requiring methods to overcome both 1) overfitting towards poorly annotated data and 2) distribution shift across domains. Unfortunately, a simple combination of domain adaptation (DA) and semi-supervised learning (SSL) methods often fail to address such two objects because of training data bias towards labeled samples. In this paper, we introduce an adaptive structure learning method to regularize the cooperation of SSL and DA. Inspired by the multi-views learning, our proposed framework is composed of a shared feature encoder network and two classifier networks, trained for contradictory purposes. Among them, one of the classifiers is applied to group target features to improve intra-class density, enlarging the gap of categorical clusters for robust representation learning. Meanwhile, the other classifier, serviced as a regularizer, attempts to scatter the source features to enhance the smoothness of the decision boundary. The iterations of target clustering and source expansion make the target features being well-enclosed inside the dilated boundary of the corresponding source points. For the joint address of cross-domain features alignment and partially labeled data learning, we apply the maximum mean discrepancy (MMD) distance minimization and self-training (ST) to project the contradictory structures into a shared view to make the reliable final decision. The experimental results over the standard SSDA benchmarks, including DomainNet and Office-home, demonstrate both the accuracy and robustness of our method over the state-of-the-art approaches.

Dynamic order Markov model for categorical sequence clustering

Markov models are extensively used for categorical sequence clustering and classification due to their inherent ability to capture complex chronological dependencies hidden in sequential data. Existing Markov models are based on an implicit assumption that the probability of the next state depends on the preceding context/pattern which is consist of consecutive states. This restriction hampers the models since some patterns, disrupted by noise, may be not frequent enough in a consecutive form, but frequent in a sparse form, which can not make use of the information hidden in the sequential data. A sparse pattern corresponds to a pattern in which one or some of the state(s) between the first and last one in the pattern is/are replaced by wildcard(s) that can be matched by a subset of values in the state set. In this paper, we propose a new model that generalizes the conventional Markov approach making it capable of dealing with the sparse pattern and handling the length of the sparse patterns adaptively, i.e. allowing variable length pattern with variable wildcards. The model, named Dynamic order Markov model (DOMM), allows deriving a new similarity measure between a sequence and a set of sequences/cluster. DOMM builds a sparse pattern from sub-frequent patterns that contain significant statistical information veiled by the noise. To implement DOMM, we propose a sparse pattern detector (SPD) based on the probability suffix tree (PST) capable of discovering both sparse and consecutive patterns, and then we develop a divisive clustering algorithm, named DMSC, for Dynamic order Markov model for categorical sequence clustering. Experimental results on real-world datasets demonstrate the promising performance of the proposed model.

Context-Based Geodesic Dissimilarity Measure for Clustering Categorical Data

Measuring the dissimilarity between two observations is the basis of many data mining and machine learning algorithms, and its effectiveness has a significant impact on learning outcomes. The dissimilarity or distance computation has been a manageable problem for continuous data because many numerical operations can be successfully applied. However, unlike continuous data, defining a dissimilarity between pairs of observations with categorical variables is not straightforward. This study proposes a new method to measure the dissimilarity between two categorical observations, called a context-based geodesic dissimilarity measure, for the categorical data clustering problem. The proposed method considers the relationships between categorical variables and discovers the implicit topological structures in categorical data. In other words, it can effectively reflect the nonlinear patterns of arbitrarily shaped categorical data clusters. Our experimental results confirm that the proposed measure that considers both nonlinear data patterns and relationships among the categorical variables yields better clustering performance than other distance measures.

Cloud deployment of game theoretic categorical clustering using apache spark: An application to car recommendation

Personal vehicles are invariably being preferred over public transport nowadays. Contact-less feature inspection and analysis based on personal preferences will be in high demand among customers in the post-pandemic world. A comprehensive online car recommendation system will be the customers’ spontaneous choice to understand and select the features of vehicles. However, the clustering of such categorical features is a challenging task as it is difficult to compare two textual attributes. In this paper, we have designed a cloud-based system that will automatically address this issue. Motivated by the cooperative game theory and fuzzy technique, and integrating the concept of Shapley theorem, a categorical data clustering algorithm has been developed. At the same time, to overcome the major limitation of having a high time complexity of the order O(n2) associated with the Shapley computation, the proposed algorithm has been distributed using Apache Spark’s Map Reduce architecture in Google Cloud Platform. The model has been thoroughly validated based on its performance on several synthetic as well as real data sets. Finally, a car recommendation system has been proposed and tested on three car sell data sets. The proposed approach outperforms the corresponding existing categorical clustering approaches in terms of various clustering validity indices. To the best of the authors’ knowledge, this is the first attempt to apply Map Reduce based Shapley computation over the categorical clustering, which can find its application beyond the proposed car recommendation system as well.

Animate and Inanimate Words Demonstrate Equivalent Retrieval Dynamics Despite the Occurrence of the Animacy Advantage.

People demonstrate a memory advantage for animate (living) concepts over inanimate (nonliving) concepts in a variety of memory tasks, including free recall, but we do not know the mechanism(s) that produces this effect. We compared the retrieval dynamics (serial-position effects, probability of first recall, output order, categorical clustering, and recall contiguity) of animate and inanimate words in a typical free recall task to help elucidate this effect. Participants were more likely to recall animate than inanimate words, but we found few, if any, differences in retrieval dynamics by word type. The animacy advantage was obtained across serial position, including occurring in both the primacy and recency regions of the lists. Participants were equally likely to recall an animate or inanimate word first on the tests and did not prioritize recalling words of one type earlier in retrieval or demonstrate strong clustering by animacy at recall. Participants showed some greater contiguity of recall for inanimate words, but this outcome ran counter to the animacy effect. Together, the results suggest that the animacy advantage stems from increased item-specific memory strength for animate over inanimate words and is unlikely to stem from intentional or strategic differences in encoding or retrieval by word type, categorical strategies, or differences in temporal organization. Although the present results do not directly support or refute any current explanations for the animacy advantage, we suggest that measures of retrieval dynamics can help to inspire or constrain future accounts for this effect and can be incorporated into relevant hypothesis testing.