Brief Announcement: Streaming Balanced Clustering

For many specific applications in data mining and machine learning, we face explicit or latent size constraint for each cluster that leads to the "balanced clustering" problem. Many existing clustering algorithms perform well in partitioning but fail in producing balanced clusters and preserving the naturally balanced structure of some data. In this paper, we propose a novel balanced clustering framework that flexibly utilizes local and global information of data. First, we propose the global balanced clustering (GBC), in which a global discriminative partitioning model is combined with the minimization of the distribution entropy of data. Then, we show that the proposed GBC can be further used to globally regularize some widely used local clustering models, so as to transform them into balanced clustering that simultaneously capture local and global data. We apply our global balanced regularization to spectral clustering (SC) and local learning (LL)-based clustering, respectively, and propose another two novel balanced clustering models: the local and global balanced SC (LGB-SC) and LGB-LL. Finding the optimal balanced partition is nondeterministic polynomial-time (NP)-hard in general. We adopt the method of augmented Lagrange multipliers to help optimize our model. Comprehensive experiments on several real world benchmarks demonstrate the advantage of our framework to yield balanced clusters while preserving good clustering quality. Our proposed LGB-SC and LGB-LL also outperform SC and LL as well as other classical clustering methods.

Scalable semantic-enhanced supervised hashing for cross-modal retrieval

We discuss several axiomatizations of entropy and some of its generalizations using an algebraic approach. The entropy is defined for such objects as functions, partitions, and set collections and the axiomatizations use natural operations on such objects. The generalizations of entropy we propose have applications in circuit design, data mining, machine learning, and information retrieval.

The diffusion of automated information retrieval (IR) ap plications and of applications which share many of the peculiarities of IR applications is increasing. The demand for these types of applications draws attention to the lack of methodologies available for IR data design. The data modell ing of an IR application, that is, the data modelling of the IR part of a complete information system, has not yet been approached and studied as a complete process and in a struc tured way. The first part of this paper introduces the motivations and the scope of the DIRD (Design of Information Retrieval Data) project, which is devoted to the development of an environ ment for the design of advanced applications of IR. The second part of the paper addresses the design of the IR part (or IR application) of an information system. The concep tual paradigm necessary for the design of advanced IR applica tions is investigated. The Entity Relationship (ER) approach is compared with that conceptual modelling paradigm and is examined as a candidate data model for the conceptual design of IR data. A new ER approach is then introduced: this new approach extends the constructs of the ER model to manage the complexity of IR data. Two design examples are given to present the use of the new ER approach in designing IR data.

With the rapid growth of the World Wide Web and the capacity of digital data storage, tremendous amount of data are generated daily from business and engineering to the Internet and science. The Internet, financial realtime data, hyperspectral imagery, and DNA microarrays are just a few of the common sources that feed torrential streams of data into scientific and business databases worldwide. Compared to statistical data sets with small size and low dimensionality, traditional clustering techniques are challenged by such unprecedented high volume, high dimensionality complex data. To meet these challenges, many new clustering algorithms have been proposed in the area of data mining (Han & Kambr, 2001). Spectral techniques have proven useful and effective in a variety of data mining and information retrieval applications where massive amount of real-life data is available (Deerwester et al., 1990; Kleinberg, 1998; Lawrence et al., 1999; Azar et al., 2001). In recent years, a class of promising and increasingly popular approaches — spectral methods — has been proposed in the context of clustering task (Shi & Malik, 2000; Kannan et al., 2000; Meila & Shi, 2001; Ng et al., 2001). Spectral methods have the following reasons to be an attractive approach to clustering problem: • Spectral approaches to the clustering problem offer the potential for dramatic improvements in efficiency and accuracy relative to traditional iterative or greedy algorithms. They do not intrinsically suffer from the problem of local optima. • Numerical methods for spectral computations are extremely mature and well understood, allowing clustering algorithms to benefit from a long history of implementation efficiencies in other fields (Golub & Loan, 1996). • Components in spectral methods have the naturally close relationship with graphs (Chung, 1997). This characteristic provides an intuitive and semantic understanding of elements in spectral methods. It Spectral Methods for Data Clustering Wenyuan Li Nanyang Technological University, Singapore Wee Keong Ng Nanyang Technological University, Singapore is important when the data is graph-based, such as links of WWW, or can be converted to graphs. In this paper, we systematically discuss applications of spectral methods to data clustering.

Document classification has important applications in information retrieval, data mining, and natural language processing and it involves categorizing documents into predefined classes. This survey paper offers a comprehensive overview of state-of-the-art techniques in document classification. Document classification encompasses of traditional machine learning algorithms like logistic regression and decision trees, alongside deep learning models such as convolutional neural networks (CNNs) and recurrent neural networks (RNNs). Furthermore, the paper highlights recent advancements in transfer learning, multi-label classification, and domain adaptation, underscoring their significance in addressing challenges such as data scarcity and domain shift. By presenting these diverse approaches, this survey aims to provide researchers and practitioners with a comprehensive understanding of various document classification techniques, paving the way for future advancements in the field.

Maximizing monotone submodular functions under a matroid constraint is a classic algorithmic problem with multiple applications in data mining and machine learning. We study this significant problem in the fully dynamic setting, where elements can be both inserted and deleted in real-time. Our main result is a randomized algorithm that maintains an efficient data structure with an \({\tilde{O}(\frac{{k^{2}}}{{\varepsilon}})}\) amortized update time (in the number of insertions and deletions) and yields a \({(4+O(\varepsilon))}\) -approximate solution with respect to the dynamic optimum, where \(k\) is the rank of the matroid.

Graph clustering is a fundamental computational problem with a number of applications in algorithm design, machine learning, data mining, and analysis of social networks. Over the past decades, researchers have proposed a number of algorithmic design methods for graph clustering. Most of these methods, however, are based on complicated spectral techniques or convex optimisation and cannot be directly applied for clustering many networks that occur in practice, whose information is often collected on different sites. Designing a simple and distributed clustering algorithm is of great interest and has comprehensive applications for processing big datasets. In this article, we present a simple and distributed algorithm for graph clustering: For a wide class of graphs that are characterised by a strong cluster-structure, our algorithm finishes in a poly-logarithmic number of rounds and recovers a partition of the graph close to optimal. One of the main procedures behind our algorithm is a sampling scheme that, given a dense graph as input, produces a sparse subgraph that provably preserves the cluster-structure of the input. Compared with previous sparsification algorithms that require Laplacian solvers or involve combinatorial constructions, this procedure is easy to implement in a distributed setting and runs fast in practice.

Integration of computing systems and methods in legal activity allows to extract benefits such as resource saving, increase objectivity, completeness and accuracy of intellectual results. Understanding of the main scientific achievements and trends at the intersection of computer and legal sciences focuses on promising scientific and technological areas of informatization of law. This review systematizes the most important achievements at the intersection of legal and computer sciences, covering foreign and domestic scientific publications for 1949–2020. Researches of computer methods and systems initiated by American jurimetrics were aimed at storing, indexing, abstracting and searching for legal texts and led to the creation of legal information retrieval systems. Domestic legal cybernetics became a pioneer in the field of automation of criminalistics expert examination. Computer modeling of legal reasoning developed on the techniques of artificial intelligence, later it shifted to the field of legal dialogue and conflict of legal arguments, but in recent years it transforms the acquired experience on the basis of modern computer models, patterns and architectures. The popularization of decision-making support systems has provided multi-tasking systems, including information retrieval, legal reasoning, analytics, predictions and control. The latest research segments are focused on the application of machine learning and big data processing. Almost all successful methodological solutions retain their significance, continuing to be applied directly or as the basis for the further development of computational methods and information systems in legal activity.

Humans have used electronic information retrieval (IR) systems for more than 50 years as they evolved from experimental systems to full-scale Web search engines and digital libraries. The fields of library and information science (LIS), cognitive science, human factors and computer science have historically been the leading disciplines in conducting research that seeks to model human interaction with IR systems for all kinds of information related behaviors. As technology problems have been mastered, the theoretical and applied framework for studying human interaction with IR systems has evolved from systems-centered to more user-centered, or cognitive-centered approaches. However, cognitive information retrieval (CIR) research that focuses on user interaction with IR systems is still largely under-funded and is often not included at computing and systems design oriented conferences. But CIR-focused research continues, and there are signs that some IR systems designers in academia and the Web search business are realizing that user behavior research can provide valuable insights into systems design and evaluation. The goal of our book is to provide an overview of new CIR research directions. This book does not provide a history of the research field of CIR. Instead, the book confronts new ways of looking at the human information condition with regard to our increasing need to interact with IR systems. The need has grown due to a number of factors, including the increased importance of information to more people in this information age. Also, IR was once considered document-oriented, but has now evolved to include multimedia, text, and other information objects. As a result, IR systems and their complexity have proliferated as users and user purposes for using them have also proliferated. Human interaction with IR systems can often be frustrating as people often lack an understanding of IR system functionality. New more holistic directions in CIR are emerging that conceptualize human-IR system interaction at the human-computer interaction level (taking into account interface issues), the information behavior level (taking into account the role of IR system interaction in the total range of people’s information behaviors), and the

Clustering plays an important role in data mining, being useful in many fields that deal with exploratory data analysis, such as information retrieval, document extraction, and image segmentation. Although they are essential in data mining applications, most clustering algorithms are adhoc methods. They have a lack of guarantee on the solution quality, which in many cases is related to a premature convergence to a local minimum of the search space. In this research, we address the problem of data clustering from an optimization perspective, where we propose a hybrid genetic algorithm to solve the Minimum Sum-of-Squares Clustering (MSSC) problem. This meta-heuristic is capable of escaping from local minima and generating near-optimal solutions to the MSSC problem. Results show that the proposed method outperformed the best current literature resultsin terms of solution quality -for almost all considered sets of benchmark instances for the MSSC objective.

Spectral clustering has become one of the most popular clustering algorithms in recent years. In real-world clustering problems, the data points for clustering may have considerable noise. To the best of our knowledge, no single clustering algorithm is able to identify all different types of cluster structures. In the existing spectral clustering methods, little effort has been made to explicitly handle both the possibly considerable noise in data points and the robustness of clustering methods, which often degrades the clustering performance. In this paper, motivated by resampling and matrix aggregation, we propose a method for robust spectral clustering. In our method, we first construct multiple transition probability matrices, each is constructed by a subset of randomly selected features. Then, these matrices can be used to recover a shared low-rank similarity matrix, which is the input to the spectral clustering, and several sparse matrices, which represent the noise. The corresponding optimization problem has a low-rank constraint on the transition probability matrix. To solve the corresponding optimization problem, an optimization procedure based on the scheme of Augmented Lagrangian Method of Multipliers is designed. Experimental results on several real-world datasets show that our method has superior performance over several state-of-the-art clustering methods.

Text segmentation is a fundamental problem in natural language processing, which has application in information retrieval, question answering, and text summarization. Almost previous works on unsupervised text segmentation are based on the assumption of lexical cohesion, which is indicated by relations between words in the two units of text. However, they only take into account the reiteration, which is a category of lexical cohesion, such as word repetition, synonym or superordinate. In this research, we investigate the non-systematic semantic relation, which is classified as collocation in lexical cohesion. This relation holds between two words or phrases in a discourse when they pertain to a particular theme or topic. This relation has been recognized via a topic model, which is, in turn, acquired from a large collection of texts. The experimental results on the public dataset show the advantages of our approach in comparison to the available unsupervised approaches.

Capacitated Clustering Problem (CCP) assigns objects to capacitated clusters. Among various CCPs, this work focuses on the capacitated P-median problem (CPMP). Imposing a strict capacity on each cluster, often leads to inefficient overlap for geographical data, which significantly diminish efficiency. We propose a four-stage clustering algorithm, applying convex polygon boundary perturbation to minimize overlap while satisfying capacity constraints. Necessitating a fixed number of clusters in many applications introduces a challenge in meeting capacity constraint. While most methods open a new cluster for points left unassigned, this paper designs a placement algorithm without allowing new clusters. The proposed algorithm demonstrates superior performance compared to three state-of-the-art approaches on specific instances, achieves similar performance on others, with some results reaching optimality. A novel metric is introduced to compare different methods on the same benchmarks.

PageRank has numerous applications in information retrieval, reputation systems, machine learning, and graph partitioning. In this paper, we study PageRank in undirected random graphs with expansion property. The Chung-Lu random graph represents an example of such graphs. We show that in the limit, as the size of the graph goes to infinity, PageRank can be represented by a mixture of the restart distribution and the vertex degree distribution.