Search

Rare efforts have been devoted to generating the structured Navigation Box (Navbox) for Wikipedia articles. A Navbox is a table in Wikipedia article page that provides a consistent navigation system for related entities. Navbox is critical for the readership and editing efficiency of Wikipedia. In this paper, we target on the automatic generation of Navbox for Wikipedia articles. Instead of performing information extraction over unstructured natural language text directly, an alternative avenue is explored by focusing on a rich set of semi-structured data in Wikipedia articles: linked entities. The core idea of this paper is as follows: If we cluster the linked entities and interpret them appropriately, we can construct a high-quality Navbox for the article entity. We propose a clustering-then-labeling algorithm to realize the idea. Experiments show that the proposed solutions are effective. Ultimately, our approach enriches Wikipedia with 1.95 million new Navboxes of high quality.

Spectral clustering is one of the most prominent clustering approaches. However, it is highly sensitive to noisy input data. In this work, we propose a robust spectral clustering technique able to handle such scenarios. To achieve this goal, we propose a sparse and latent decomposition of the similarity graph used in spectral clustering. In our model, we jointly learn the spectral embedding as well as the corrupted data - thus, enhancing the clustering performance overall. We propose algorithmic solutions to all three established variants of spectral clustering, each showing linear complexity in the number of edges. Our experimental analysis confirms the significant potential of our approach for robust spectral clustering. Supplementary material is available at www.kdd.in.tum.de/RSC.

The analysis of flow traces can help to understand a network's usage patterns. We present a hierarchical clustering algorithm for network flow data that can summarize terabytes of IP traffic into a parsimonious tree model. The method automatically finds an appropriate scale of aggregation so that each cluster represents a local maximum of the traffic density from a block of source addresses to a block of destination addresses. We apply this clustering method on NetFlow data from an enterprise network, find the largest traffic clusters, and analyze their stationarity across time. The existence of heavy-volume clusters that persist over long time scales can help network operators to perform usage-based accounting, capacity provisioning and traffic engineering. Also, changes in the layout of hierarchical clusters can facilitate the detection of anomalies and significant changes in the network workload.

Document clustering is an attractive field that interests increasingly the research community, and so giving rise to several clustering algorithms. In addition to this, document collections are expanding continuously which limits the traditional and sequential clustering methods. In this paper, we present a new parallel design of a recently appeared heuristic for unsupervised and hard clustering. The original method finds clusters by partitioning categorical large data sets according to a linear mathematical model known as the Relational Analysis. We have improved this method to handle document clustering using the MapReduce programming model iteratively. The efficiency of the proposed method is evaluated using the BBC data set provided by the BBC News website. And the experiments have proved that this method yields quality results with low computational cost.

Clustering text and link data, as an important task in text and link analysis, aims at finding communities of linked documents by leveraging the information from both domains. Due to its improved performance over the single domain counterpart, it has attracted increasing attention from practitioners in recent years. Despite its popularity, all existing algorithms on clustering text and link data overlook the existence of domain-specific distinctions and thus result in unsatisfactory clustering quality. In this paper, we address this limitation by explicitly modeling the domain-specific distinctions in the clustering process. Specifically, we extend the idea of consensus and domain-specific subspace decomposition from flat data to graph data. Such a modeling, when coupled with a regularization to further sharpen the information distinction, makes the consensus information between text and link more accurate for clustering with both domains. The final model is cast into the spectral clustering model by imposing the subspace orthogonality. To eschew the costly eigen-decomposition required for spectral clustering and further speed-up the optimization, we take advantage of the data sparsity and the low dimensionality of subspaces, and deploy a constraint-preserving gradient method to efficiently solve the model. The experimental study on three real datasets shows that our algorithm consistently and significantly outperforms the state-of-the-art relevant algorithms in terms of both quality and efficiency.

International audience; Clustering aims to classify elements into groups called classes or clusters. Clustering is used in reverse-engineering to help to understand legacy software. It is also a tech-nic used in re-engineering to propose gatherings of software entities to engineers who can then accept them or not. This paper presents a Pharo implementation of an iterative and semi-automatic method for clustering. Our method proposes, to an end-user, clusters that are based on domain information and structural information. The method presented in this paper has been applied in an industrial project of architecture migration. We show that this method helps engineers to cluster software elements into domain concepts. The clustering gives a result of 56% of precision and 79% of recall after the automated part in a high level clustering. A deeper clustering gives a result of 51% of precision and 52% of recall.

Clustering is widely used in a variety of fields to find structures among data and extract useful knowledge. Recently, there has been an emergent need for robust and efficient techniques that can manage the exploding volume of data available in the World Wide Web or gathered from devices and sensors. However, clustering such data is challenging, due to the multimodal nature of this information. In this work, we introduce a novel Multimodal Adaptive Genetic Clustering (MAGC) algorithm that clusters information based on multiple features. Our approach adds feature weights as an extension to the chromosome, which represents a clustering solution, such that feature weights are also evolved and optimized alongside the original clustering solution. The number of clusters is also adaptive and is optimized during the search.

An efficient and effective clustering process is a core task of data mining analysis, and has become more important in the nowadays scenario of big data, where scalability is an issue. In this paper we present the ClusMAM method, which proposes a new strategy for clustering large complex datasets through metric access methods. ClusMAM aims at accelerating the process of relational partitional clustering by taking advantage of the inherent node separations of metric access methods. In comparison with other methods from the literature, ClusMAM is up to four orders of magnitude faster than the competitors maintaining clustering quality. Additionally, ClusMAM exploits the datasets to find compact and coherent clusters, suggesting the number of clusters k found in the data. The method was evaluated employing synthetic and real datasets, and the behavior of the method was consistent regarding the number of distance calculations and time required for the clustering process as well.

A relatively new feature in Google Play Store presents mobile app search results grouped by topic, helping users to quickly navigate and explore. The underlying Search Results Clustering (SRC) system faces several challenges, including grouping search results in topical coherent clusters as well as finding the appropriate level of granularity for clustering. We present AppGrouper, an alternative approach to algorithmic-only solutions, incorporating human input in a knowledge-graph-based clustering process. AppGrouper provides an interactive interface that lets domain experts steer the clustering process in early, mid, and late stages. We deployed and evaluated AppGrouper with internal experts. We found that AppGroup improved quality of algorithm-generated app clusters on 56 out of 82 search queries. We also found that the internal experts made more changes in early and mid stages for lower-quality algorithmic results, focusing more on narrow queries. Our result suggests, in some contexts, machine learning systems can greatly benefit from steering from human experts, creating a symbiotic working relationship.

Multi-view clustering, which explores complementary information between multiple distinct feature sets for better clustering, has a wide range of applications, e.g., knowledge management and information retrieval. Traditional multi-view clustering methods usually assume that all examples have complete feature sets. However, in real applications, it is often the case that some examples lose some feature sets, which results in incomplete multi-view data and notable performance degeneration. In this paper, a novel incomplete multi-view clustering method is therefore developed, which learns unified latent representations and projection matrices for the incomplete multi-view data. To approximate the high level scaled indicator matrix defined to represent class label matrix, the latent representations are expected to be non-negative and column orthogonal. Besides, since data are often with high dimensional and noisy features, the projection matrices are enforced to be sparse so as to select relevant features when learning the latent space. Furthermore, the inter-view and intra-view data structure is preserved to further enhance the clustering performance. To these ends, an objective is developed with efficient optimization strategy and convergence analysis. Extensive experiments demonstrate that our model performs better than the state-of-the-art multi-view clustering methods in various settings.

Many commonly used data-mining techniques utilized across research fields perform poorly when used for large data sets. Sequential agglomerative hierarchical non-overlapping clustering is one technique for which the algorithms’ scaling properties prohibit clustering of a large amount of items. Besides the unfavorable time complexity of O(n 2 ), these algorithms have a space complexity of O(n 2 ), which can be reduced to O(n) if the time complexity is allowed to rise to O(n 2 log2 n). In this paper, we propose the use of locality-sensitive hashing combined with a novel data structure called twister tries to provide an approximate clustering for average linkage. Our approach requires only linear space. Furthermore, its time complexity is linear in the number of items to be clustered, making it feasible to apply it on a larger scale. We evaluate the approach both analytically and by applying it to several data sets. peerReviewed

The incorporation of semi-supervision in the cluster detection process has proved especially useful when one wants to get a high consistency between the data partitioning and the knowledge the user has about the data domain. In recent years, several strategies for semi-supervised clustering have been proposed. The approaches adopted by these strategies aim at guiding the process of cluster detection by using constraints: to interfere with the allocation of elements to the most appropriate cluster at each iteration of the algorithm; or to modify the objective function employed. This paper proposes a novel approach for incorporating semi-supervision in the well-known k-means algorithm. This semi-supervised clustering method employs constraint information in the definition of multiple assistant representatives for the centroids used at each iteration of k-means. A refinement process is designed to reduce the number of assistant representatives considered for each centroid without losing the clustering quality. The experimental results with eight synthetic datasets show the potential of the proposed approach for dealing with complex data structures composed by clusters of different shapes.

Clustering high-dimensional data spaces are often encountered in areas such as medicine, DNA analysis in computational biology and many others. It imposes on a data analysis severe computational requirements and present real challenges to clustering algorithms. This paper evaluates the performance efficiency of K-means and Agglomerative hierarchical clustering methods based on Euclidean and Manhattan distance functions for high dimensional data. Efficiency concerns the computational time required to build up datasets. Extensive experiments carried out to evaluate two clustering methods on Microarray datasets. The results demonstrate that Agglomerative hierarchical clustering algorithm is efficient in time for both distance functions than K-means clustering algorithm.

Information Retrieval (IR) systems such as search engines retrieve a large set of documents, images and videos in response to a user query. Computational methods such as Automatic Text Summarization (ATS) reduce this information load enabling users to find information quickly without reading the original text. The challenges to ATS include both the time complexity and the accuracy of summarization. Our proposed Information Retrieval system consists of three different phases: Retrieval phase, Clustering phase and Summarization phase. In the Clustering phase, we extend the Potential-based Hierarchical Agglomerative (PHA) clustering method to a hybrid PHA-ClusteringGain-K-Means clustering approach. Our studies using the DUC 2002 dataset show an increase in both the efficiency and accuracy of clusters when compared to both the conventional Hierarchical Agglomerative Clustering (HAC) algorithm and PHA.

Biodiversity assessment is an important step in a metagenomic processing pipeline. The biodiversity of a microbial metagenome is often estimated by grouping its 16S rRNA reads into operational taxonomic units or OTUs. These metagenomic datasets are typically large and hence require effective yet accurate computational methods for processing.In this paper, we introduce a new hierarchical clustering method called CRiSPy-Embed which aims to produce high-quality clustering results at a low computational cost. We tackle two computational issues of the current OTU hierarchical clustering approach: (1) the compute-intensive sequence alignment operation for building the distance matrix and (2) the quadratic memory requirement of the clustering procedure.Our performance evaluation shows that CRiSPy-Embed achieves higher efficiency in terms of both runtime and memory consumption in comparison to existing dendrogram-based approaches. Furthermore, to obtain the final OTU grouping, CRiSPy-Embed dynamically determines a natural cutoff of the dendrogram. With this strategy, CRiSPy-Embed achieves better and more robust clustering outcomes compared to other notable OTU clustering pipelines.

Correlation clustering is arguably the most natural formulation of clustering. Given a set of objects and a pairwise similarity measure between them, the goal is to cluster the objects so that, to the best possible extent, similar objects are put in the same cluster and dissimilar objects are put in different clusters. As it just needs a definition of similarity, its broad generality makes it applicable to a wide range of problems in different contexts, and in particular makes it naturally suitable to clustering structured objects for which feature vectors can be difficult to obtain.Despite its simplicity, generality and wide applicability, correlation clustering has so far received much more attention from the algorithmic theory community than from the data mining community. The goal of this tutorial is to show how correlation clustering can be a powerful addition to the toolkit of the data mining researcher and practitioner, and to encourage discussions and further research in the area.In the tutorial we will survey the problem and its most common variants, with an emphasis on the algorithmic techniques and key ideas developed to derive efficient solutions. We will motivate the problems and discuss real-world applications, the scalability issues that may arise, and the existing approaches to handle them.

Short text clustering has become an increasingly important task with the popularity of social media like Twitter, Google+, and Facebook. It is a challenging problem due to its sparse, high-dimensional, and large-volume characteristics. In this paper, we proposed a collapsed Gibbs Sampling algorithm for the Dirichlet Multinomial Mixture model for short text clustering (abbr. to GSDMM). We found that GSDMM can infer the number of clusters automatically with a good balance between the completeness and homogeneity of the clustering results, and is fast to converge. GSDMM can also cope with the sparse and high-dimensional problem of short texts, and can obtain the representative words of each cluster. Our extensive experimental study shows that GSDMM can achieve significantly better performance than three other clustering models.

Different important studies in Web search results clustering have recently shown increasing performances motivated by the use of external resources. Following this trend, we present a new algorithm called Dual C-Means, which provides a theoretical background for clustering in different representation spaces. Its originality relies on the fact that external resources can drive the clustering process as well as the labeling task in a single step. To validate our hypotheses, a series of experiments are conducted over different standard datasets and in particular over a new dataset built from the TREC Web Track 2012 to take into account query logs information. The comprehensive empirical evaluation of the proposed approach demonstrates its significant advantages over traditional clustering and labeling techniques.

Typically, recommendations are computed by considering users similar to the user in question. However, scanning the whole database of users for locating similar users is expensive. Existing approaches build user profiles by employing full-dimensional clustering to find sets of similar users. As the datasets we deal with are high-dimensional and incomplete, full-dimensional clustering is not the best option. To this end, we explore the fault tolerance subspace clustering approach that detects clusters of similar users in subspaces of the original feature space and also allows for missing values. Our experiments on real movie datasets show that the diversification of the similar users through subspace clustering results in better recommendations comparing to traditional collaborative filtering and full dimensional clustering approaches.

Spatial clustering deals with the unsupervised grouping of places into clusters and finds important applications in urban planning and marketing. Current spatial clustering models disregard information about the people who are related to the clustered places. In this paper, we show how the density-based clustering paradigm can be extended to apply on places which are visited by users of a geo-social network. Our model considers both spatial information and the social relationships between users who visit the clustered places. After formally defining the model and the distance measure it relies on, we present efficient algorithms for its implementation, based on spatial indexing. We evaluate the effectiveness of our model via a case study on real data; in addition, we design two quantitative measures, called social entropy and community score to evaluate the quality of the discovered clusters. The results show that geo-social clusters have special properties and cannot be found by applying simple spatial clustering approaches. The efficiency of our index-based implementation is also evaluated experimentally.

We study the problem of learning personalized user models from rich user interactions. In particular, we focus on learning from clustering feedback (i.e., grouping recommended items into clusters), which enables users to express similarity or redundancy between different items. We propose and study a new machine learning problem for personalization, which we call collaborative clustering. Analogous to collaborative filtering, in collaborative clustering the goal is to leverage how existing users cluster or group items in order to predict similarity models for other users' clustering tasks. We propose a simple yet effective latent factor model to learn the variability of similarity functions across a user population. We empirically evaluate our approach using data collected from a clustering interface we developed for a goal-oriented data exploration (or sensemaking) task: asking users to explore and organize attractions in Paris. We evaluate using several realistic use cases, and show that our approach learns more effective user models than conventional clustering and metric learning approaches.

Clustering Web 2.0 items (i.e., web resources like videos, images) into semantic groups benefits many applications, such as organizing items, generating meaningful tags and improving web search. In this paper, we systematically investigate how user-generated comments can be used to improve the clustering of Web 2.0 items. In our preliminary study of Last.fm, we find that the two data sources extracted from user comments -- the textual comments and the commenting users -- provide complementary evidence to the items' intrinsic features. These sources have varying levels of quality, but we importantly we find that incorporating all three sources improves clustering. To accommodate such quality imbalance, we invoke multi-view clustering, in which each data source represents a view, aiming to best leverage the utility of different views. To combine multiple views under a principled framework, we propose CoNMF (Co-regularized Non-negative Matrix Factorization), which extends NMF for multi-view clustering by jointly factorizing the multiple matrices through co-regularization. Under our CoNMF framework, we devise two paradigms -- pair-wise CoNMF and cluster-wise CoNMF -- and propose iterative algorithms for their joint factorization. Experimental results on Last.fm and Yelp datasets demonstrate the effectiveness of our solution. In Last.fm, CoNMF betters k-means with a statistically significant F1 increase of 14%, while achieving comparable performance with the state-of-the-art multi-view clustering method CoSC (Co-regularized Spectral Clustering). On a Yelp dataset, CoNMF outperforms the best baseline CoSC with a statistically significant performance gain of 7%.

Efficient techniques for face clustering, along with speech and text recognition, can provide the means for rapid browsing and accurate retrieval of videos from large video databases. Observing that video data contains information about not only the face features, but also the temporal ordering of the faces and positional coordinates of the face-regions within the video frame, an attempt is made to consolidate all of these details into the framework of the flexible and intuitive hierarchical clustering algorithm. This paper outlines a novel initialization mechanism for the hierarchical clustering to follow, based on the temporal and positional information of the face-samples extracted from a video. Experiments with news broadcast videos show that the novel algorithm is considerably more efficient than it's parent, and is promising for future exploration.

Finding the appropriate information and understanding to human research is a delicate task when dealing with an outstanding number of unstructured texts created daily. Hence the objective of clustering algorithms which are part of the powerful text mining tools. In this paper, we propose a novel text document clustering based on a new hybrid feature selection method that we call HFSM. This technique extracts statistical and semantic relevant terms to pilot the clustering mechanism. The experiments conducted on Reuters corpus demonstrate the practical aspects of our algorithm and show that it generates more accurate clustering than the one obtained by other existing algorithms.

Automated clustering of threads within and across web forums will greatly benefit both users and forum administrators in efficiently seeking, managing, and integrating the huge volume of content being generated. While clustering has been studied for other types of data, little work has been done on clustering forum threads; the informal nature and special structure of forum data make it interesting to study how to effectively cluster forum threads. In this paper, we apply three state of the art clustering methods (i.e., hierarchical agglomerative clustering, k-Means, and probabilistic latent semantic analysis) to cluster forum threads and study how to leverage the structure of threads to improve clustering accuracy. We propose three different methods for assigning weights to the posts in a forum thread to achieve more accurate representation of a thread. We evaluate all the methods on data collected from three different Linux forums for both within-forum and across-forum clustering. Our results show that the state of the art methods perform reasonably well for this task, but the performance can be further improved by exploiting thread structures. In particular, a parabolic weighting method that assigns higher weights for both beginning posts and end posts of a thread is shown to consistently outperform a standard clustering method.

This paper addresses the problem of learning to classify texts by exploiting information derived from clustering both training and testing sets. The incorporation of knowledge resulting from clustering into the feature space representation of the texts is expected to boost the performance of a classifier. Two different approaches to clustering are described, an unsupervised and a semi-supervised one. We present an empirical study of the proposed algorithms on a variety of datasets. The results are encouraging, revealing that information resulting from clustering can create text classifiers of high-accuracy.

In many text clustering tasks, there is some valuable knowledge about the problem domain, in addition to the original textual data involved in the clustering process. Traditional text clustering methods are unable to incorporate such additional (privileged) information into data clustering. Recently, a new paradigm called LUPI - Learning Using Privileged Information - was proposed by Vapnik to incorporate privileged information in classification tasks. In this paper, we extend the LUPI paradigm to deal with text clustering tasks. In particular, we show that the LUPI paradigm is potentially promising for incremental hierarchical text clustering, being very useful for organizing large textual databases. In our method, the privileged information about the text documents is applied to refine an initial clustering model by means of consensus clustering. The initial model is used for incremental clustering of the remaining text documents. We carried out an experimental evaluation on two benchmark text collections and the results showed that our method significantly improves the clustering accuracy when compared to a traditional hierarchical clustering method.

Numerical schemes research on clustering model has been quite intensive in the past decade. The difficulties associated with curse of dimensionality and cost functions to reflect the general knowledge about internal structures and distributions of target data. Traditional computational clustering and variables selection schemes are struggling to estimate at high level of accuracy for this type of problem. Hence, in the present study, a novel semantic-based scheme was proposed to enhance the clustering accuracy. The results show that our conceptual model is automatic and optimal. Good comparisons with the experimental studies demonstrate the multidisciplinary applications of our approach.

Finding multiple clustering solutions has recently gained much attention. Based on the observation that data is often multi-faceted, novel clustering methods have been introduced capable of detecting multiple, diverse clusterings. In this work-in-progress paper, we present a novel stochastic subspace search principle that tackles the requirements of multi-view clustering. The main idea is to consider each subspace as a state in a Markov chain and using Monte Carlo methods to sample the multi-view subspaces. By dynamically adapting the underlying probability density function we realize the generation of alternative clustering views. We present preliminary experimental results of our method and we describe future research directions.

Clustering is a basic data mining task with a wide variety of applications. Not surprisingly, there exist many clustering algorithms. However, clustering is an ill defined problem - given a data set, it is not clear what a "correct" clustering for that set is. Indeed, different algorithms may yield dramatically different outputs for the same input sets. In contrast with other common learning tasks, like classification prediction, clustering does not have a well defined ground truth. Faced with a concrete clustering task, a user needs to choose an appropriate clustering algorithm (as well as a concrete setting for the tunable parameters of the chosen algorithm). Currently, such decisions are often made in a very ad hoc, if not completely random, manner. Given the crucial effect of the choice of a clustering algorithm on the resulting clustering, this state of affairs is truly regrettable. Can the research community develop effective tools for helping users make informed decisions when they come to pick a clustering tool for their data? How can we help the data analysts in finding the cluster structures are looking for?Several paradigms have been proposed to answer that challenge. These include, semi-supervised clustering (in which the user specifies partial information about the desired clustering solution in the form of link/don't link examples) and multiple clusterings, as well as tools for visualization of clusterings. In this work, we propose a high-level approach to this challenge. The basic premise of my work is that prior domain knowledge is an indispensable component of any successful cluttering paradigm. In light of this, a major research objective is the development tools for communicating relevant prior knowledge between the data analysts, that has some understanding or intuition about the task at hand, and the tools for choosing clusterings (or clustering algorithms).We address this objective by proposing two approaches. First, we address the choice of clustering algorithm. Our paradigm is to distill abstract properties of the input-output behaviors of different clustering paradigms. The goal is to come up with a list of such properties so that these properties can capture some of the domain knowledge that users have about their tasks, while being strong enough to distinguish between different clustering algorithms. We introduce several abstract properties of clustering functions and use them to taxonomize clustering algorithmic paradigms. Secondly, we consider requirements for defining the quality of given clusterings. Such clustering quality measures can be viewed as another way for expressing prior domain knowledge.

Consensus clustering and interactive feature selection are very useful methods to extract and manage knowledge from texts. While consensus clustering allows the aggregation of different clustering solutions into a single robust clustering solution, the interactive feature selection facilitates the incorporation of the users experience in text clustering tasks by selecting a set of high-level features. In this paper, we propose an approach to improve the robustness of consensus clustering using interactive feature selection. We have reported some experimental results on real-world datasets that show the effectiveness of our approach.

Ill-structured code is difficult to understand and thereby is costly to maintain and reuse. Software restructuring techniques based on hierarchical agglomerative clustering (HAC) algorithms have been widely used to restructure large modules with low cohesion into smaller modules with high cohesion, without changing the overall behaviour of the software. These techniques generate clustering trees, of modules, that are sliced at different cut-points to obtain desired restructurings. Choosing appropriate cut-points has always been a difficult problem in clustering. Previous HAC techniques generate clustering trees that have large number of cut-points. Moreover, many of those cut-points return clusters of which only a few lead to a meaningful restructuring of the software. In this paper, we give a new hierarchical clustering technique, the (k, w)-Core Clustering ((k, w)-CC) technique, for restructuring software at the function level that generates clustering trees with lower number of cut-points, which yield a lower number of redundant clusters. (k, w)-CC gives good restructurings. To establish this, we provide an experimental comparison of (k, w)-CC with four previous HAC techniques: single linkage algorithm (SLINK), complete linkage algorithm (CLINK), weighted pair group method of arithmetic averages (WPGMA), and adaptive k-nearest neighbour algorithm (A-KNN). In the experiments, the techniques were implemented on Java functions extracted from real-life industrial programs.

We propose a competence partitioning strategy for Web search result presentation: the unmodified head of a ranked result list is combined with a clustering of documents from the result list tail. We identify two principles to which such a clustering must adhere to improve the user's search experience: (1) Avoid the unwanted effect of query aspect repetition, which is called shadowing here. (2) Avoid extreme clusterings, i.e., neither the number of cluster labels nor the number of documents per cluster should exceed the size of the result list head. We present measures to quantify the shadowing effect, and with Faceted Clustering we introduce an algorithm that optimizes the identified principles. The key idea of Faceted Clustering is a dynamic, user-controlled reorganization of a clustering, similar to a faceted navigation system. We report on evaluations using the AMBIENT corpus and demonstrate the potential of our approach by a comparison with two well-known clustering search engines.

Evolutionary data, such as topic changing blogs and evolving trading behaviors in capital market, is widely seen in business and social applications. The time factor and intrinsic change embedded in evolutionary data greatly challenge evolutionary clustering. To incorporate the time factor, existing methods mainly regard the evolutionary clustering problem as a linear combination of snapshot cost and temporal cost, and reflect the time factor through the temporal cost. It still faces accuracy and scalability challenge though promising results gotten. This paper proposes a novel evolutionary clustering approach, evolutionary maximum margin clustering (e-MMC), to cluster large-scale evolutionary data from the maximum margin perspective. e-MMC incorporates two frameworks: Data Integration from the data changing perspective and Model Integration corresponding to model adjustment to tackle the time factor and change, with an adaptive label allocation mechanism. Three e-MMC clustering algorithms are proposed based on the two frameworks. Extensive experiments are performed on synthetic data, UCI data and real-world blog data, which confirm that e-MMC outperforms the state-of-the-art clustering algorithms in terms of accuracy, computational cost and scalability. It shows that e-MMC is particularly suitable for clustering large-scale evolving data.

With the development of statistical machine translation, we have ready-to-use tools that can translate documents from one language to many other languages. These translations provide different yet correlated views of the same set of documents. This gives rise to an intriguing question: can we use the extra information to achieve a better clustering of the documents? Some recent work on multiview clustering provided positive answers to this question. In this work, we propose an alternative approach to address this problem using the constrained clustering framework. Unlike traditional Must-Link and Cannot-Link constraints, the constraints generated from machine translation are dense yet noisy. We show how to incorporate this type of constraints by presenting two algorithms, one parametric and one non-parametric. Our algorithms are easy to implement, efficient, and can consistently improve the clustering of real data, namely the Reuters RCV1/RCV2 Multilingual Dataset. In contrast to existing multiview clustering algorithms, our technique does not need the compatibility or the conditional independence assumption, nor does it involve subtle parameter tuning.

Real-world, multiple-typed objects are often interconnected, forming heterogeneous information networks. A major challenge for link-based clustering in such networks is its potential to generate many different results, carrying rather diverse semantic meanings. In order to generate desired clustering, we propose to use meta-path, a path that connects object types via a sequence of relations, to control clustering with distinct semantics. Nevertheless, it is easier for a user to provide a few examples ("seeds") than a weighted combination of sophisticated meta-paths to specify her clustering preference. Thus, we propose to integrate meta-path selection with user-guided clustering to cluster objects in networks, where a user first provides a small set of object seeds for each cluster as guidance. Then the system learns the weights for each meta-path that are consistent with the clustering result implied by the guidance, and generates clusters under the learned weights of meta-paths. A probabilistic approach is proposed to solve the problem, and an effective and efficient iterative algorithm, PathSelClus, is proposed to learn the model, where the clustering quality and the meta-path weights are mutually enhancing each other. Our experiments with several clustering tasks in two real networks demonstrate the power of the algorithm in comparison with the baselines.

Graph clustering, also known as community detection, is a long-standing problem in data mining. However, with the proliferation of rich attribute information available for objects in real-world graphs, how to leverage structural and attribute information for clustering attributed graphs becomes a new challenge. Most existing works take a distance-based approach. They proposed various distance measures to combine structural and attribute information. In this paper, we consider an alternative view and propose a model-based approach to attributed graph clustering. We develop a Bayesian probabilistic model for attributed graphs. The model provides a principled and natural framework for capturing both structural and attribute aspects of a graph, while avoiding the artificial design of a distance measure. Clustering with the proposed model can be transformed into a probabilistic inference problem, for which we devise an efficient variational algorithm. Experimental results on large real-world datasets demonstrate that our method significantly outperforms the state-of-art distance-based attributed graph clustering method.

Web Person Disambiguation is often conducted through clustering web documents to identify different namesakes for a given name. This paper presents a new key-phrased clustering method combined with a second step re-classification to identify outliers to improve cluster performance. For document clustering, the hierarchical agglomerative approach is conducted based on the vector space model which uses key phrases as the main feature. Outliers of cluster results are then identified through a centroids-based method. The outliers are then reclassified by the SVM classifier into the more appropriate clusters using a key phrase-based string kernel model as its feature space. The re-classification uses the clustering result in the first step as its training data so as to avoid the use of separate training data required by most classification algorithms. Experiments conducted on the WePS-2 dataset show that the algorithm based on key phrases is effective in improving the WPD performance.

Traditional semi-supervised clustering uses only limited user supervision in the form of labeled instances and pairwise instance constraints to aid unsupervised clustering. However, user supervision can also be provided in alternative forms for document clustering, such as labeling a feature by indicating whether it discriminates among clusters. This paper thus fills this void by enhancing traditional semi-supervised clustering with feature supervision which asks the user to label discriminating features during labeling the instance or pairwise instance constraints. Various types of semi-supervised clustering algorithms were explored with feature supervision. Our experimental results on several real-world datasets demonstrate that augmenting the instance-level supervision with feature-level supervision can significantly improve document clustering performance.

Semi-supervised clustering algorithms for general problems use a small amount of labeled instances or pairwise instance constraints to aid the unsupervised clustering. However, user supervision can also be provided in alternative forms for document clustering, such as labeling a feature by associating it with a document or a cluster. Besides labeled documents, this paper also explores labeled features to generate cluster seeds to seed the unsupervised clustering. In this paper, we present a unified framework in which one can use both labeled documents and features in terms of seeding clusters and refine this information using intermediate clusters. We introduce two methods of using labeled features to generate cluster seeds. Experimental results on several real-world data sets demonstrate that constraining the clustering by both documents and features seeding can significantly improve document clustering performance over random seeding and document only seeding.

In this paper, we address the problem of semi-supervised hierarchical clustering by using an active clustering solution with cluster-level constraints. This active learning approach is based on a concept of merge confidence in agglomerative clustering. The proposed method was compared with an un-supervised algorithm (average-link) and a semi-supervised algorithm based on pairwise constraints. The results show that our algorithm tends to be better than the pairwise constrained algorithm and can achieve a significant improvement when compared to the unsupervised algorithm.

How to organize and present search results plays a critical role in the utility of search engines. Due to the unprecedented scale of the Web and diversity of search results, the common strategy of ranked lists has become increasingly inadequate, and clustering has been considered as a promising alternative. Clustering divides a long list of disparate search results into a few topic-coherent clusters, allowing the user to quickly locate relevant results by topic navigation. While many clustering algorithms have been proposed that innovate on the automatic clustering procedure, we introduce ClusteringWiki, the first prototype and framework for personalized clustering that allows direct user editing of the clustering results. Through a Wiki interface, the user can edit and annotate the membership, structure and labels of clusters for a personalized presentation. In addition, the edits and annotations can be shared among users as a mass-collaborative way of improving search result organization and search engine utility.

The area of constrained clustering has been actively pursued for the last decade. A more recent extension that will be the focus of this paper is constrained hierarchical clustering which allows building user-constrained dendrograms/trees. Like all forms of constrained clustering, previous work on hierarchical constrained clustering uses simple constraints that are typically implemented in a procedural language. However, there exists mature results and packages in the fields of constraint satisfaction languages and solvers that the constrained clustering field has yet to explore. This work marks the first steps towards introducing constraints satisfaction languages/solvers into hierarchical constrained clustering. We make several significant contributions. We show how many existing and new constraints for hierarchical clustering, can be modeled as a Horn-SAT problem that is easily solvable in polynomial time and which allows their implementation in any number of declarative languages or efficient solvers. We implement our own solver for efficiency reasons. We then show how to formulate constrained hierarchical clustering in a flexible manner so that any number of algorithms, whose output is a dendrogram, can make use of the constraints.

How to organize and present search results plays a critical role in the utility of search engines. Due to the unprecedented scale of the Web and diversity of search results, the common strategy of ranked lists has become increasingly inadequate, and clustering has been considered as a promising alternative. Clustering divides a long list of disparate search results into a few topic-coherent clusters, allowing the user to quickly locate relevant results by topic navigation. While many clustering algorithms have been proposed that innovate on the automatic clustering procedure, we introduce ClusteringWiki, the first prototype and framework for personalized clustering that allows direct user editing of clustering results. Through a Wiki interface, the user can edit and annotate the membership, structure and labels of clusters for a personalized presentation. In addition, the edits and annotations can be shared among users as a mass collaborative way of improving search result organization and search engine utility.

Document clustering is a popular research topic, which aims to partition documents into groups of similar objects (i.e., clusters), and has been widely used in many applications such as automatic topic extraction, document organization and filtering. As a recently proposed concept, Universum is a collection of "non-examples" that do not belong to any concept/cluster of interest. This paper proposes a novel document clustering technique -- Document Clustering with Universum, which utilizes the Universum examples to improve the clustering performance. The intuition is that the Universum examples can serve as supervised information and help improve the performance of clustering, since they are known not belonging to any meaningful concepts/clusters in the target domain. In particular, a maximum margin clustering method is proposed to model both target examples and Universum examples for clustering. An extensive set of experiments is conducted to demonstrate the effectiveness and efficiency of the proposed algorithm.

Projective Clustering Ensembles (PCE) are a very recent advance in data clustering research which combines the two powerful tools of clustering ensembles and projective clustering.Specifically, PCE enables clustering ensemble methods to handle ensembles composed by projective clustering solutions. PCE has been formalized as an optimization problem with either a two-objective or a single-objective function. Two-objective PCE has shown to generally produce more accurate clustering results than its single-objective counterpart, although it can handle the object-based and feature-based cluster representations only independently of one other. Moreover, both the early formulations of PCE do not follow any of the standard approaches of clustering ensembles, namely instance-based, cluster-based, and hybrid. In this paper, we propose an alternative formulation to the PCE problem which overcomes the above issues. We investigate the drawbacks of the early formulations of PCE and define a new single-objective formulation of the problem. This formulation is capable of treating the object- and feature-based cluster representations as a whole, essentially tying them in a distance computation between a projective clustering solution and a given ensemble. We propose two cluster-based algorithms for computing approximations to the proposed PCE formulation, which have the common merit of conforming to one of the standard approaches of clustering ensembles. Experiments on benchmark datasets have shown the significance of our PCE formulation, as both the proposed heuristics outperform existing PCE methods.

We compare various document clustering techniques including K-means, SVD-based method and a graph-based approach and their performance on short text data collected from Twitter. We define a measure for evaluating the cluster error with these techniques. Observations show that graph-based approach using affinity propagation performs best in clustering short text data with minimal cluster error.

Gesture recognition is an important aspect of interpersonal social interaction. Developing a similar capacity in a robot will improve human-robot interaction. Various unsupervised clustering methods applied to clustering a set of dynamic human arm gestures are compared. Unsupervised clustering is important in gesture recognition as it imposes no a priori bound on the set of gestures. Results are compared using v-measure, a metric that allows differential weighting between clustering homogeneity and completeness. Experiments show that the best clustering method depends on the desired balance between homogeneity and completeness.

In this paper, we present a system for clustering the search results of a news search engine. The news search interface includes the relevant news articles to a given query organized in terms of related news stories. Here each cluster corresponds to a news story and the news articles are clustered into stories. We present a system that clusters the search results of a news search system in a fast and scalable manner. The clustering system is organized into three components including offline clustering, incremental clustering and realtime clustering. We propose novel techniques for clustering the search results in realtime. The experimental results with large collections of news documents reveal that our system is both scalable and also achieves good accuracy in clustering the news search results.

In this paper, for efficient clustering of visual image data that have arbitrary mixture distributions, we propose a simple distance metric learning method called Maximum Normalized Spacing (MNS) which is a generalized principle based on Maximum Spacing [12] and Minimum Spanning Tree (MST). The proposed Normalized Spacing (NS) can be viewed as a kind of adaptive distance metric for contextual dissimilarity measure which takes into account the local distribution of the data vectors. Image clustering is a difficult task because there are multiple nonlinear manifolds embedded in the data space. Many of the existing clustering methods often fail to learn the whole structure of the multiple manifolds and they are usually not very effective. Combining both the internal and external statistics of clusters to capture the density structure of manifolds, MNS is capable of efficient and effective solving the clustering problem for the complex multi-manifold datasets in arbitrary metric spaces. We apply this MNS method into the practical problem of multi-view image clustering and obtain good results which are helpful for image browsing systems. Using the COIL-20 [19] and COIL-100 [18] multi-view image databases, our experimental results demonstrate the effectiveness of the proposed MNS clustering method and this clustering method is more efficient than the traditional clustering methods.