The goal of the 24 th International Workshop on Data Mining in Bioinformatics (BIOKDD 2025) is to encourage KDD researchers to solve the numerous problems and challenges in Bioinformatics using Data Mining technologies.Based on the organizers' expertise and communities, BIOKDD 2025 features the theme "Generative AI in Biomolecular Designs".This theme encourages using large language models to design and optimize proteins or other biomolecules such as compounds, natural products, and RNAs with generative AI techniques.We also welcome broader research applying data mining to address biomedical problems.The key goal is to accelerate the convergence between Data Mining and Bioinformatics communities to expedite discoveries in basic biology, medicine and healthcare.

In the rapidly evolving landscape of online recipe sharing within a globalized context, there has been a notable surge in research towards comprehending and generating food recipes. Recent advancements in large language models (LLMs) like GPT-2 and LLaVA have paved the way for Natural Language Processing (NLP) approaches to delve deeper into various facets of food-related tasks, encompassing ingredient recognition and comprehensive recipe generation. Despite impressive performance and multi-modal adaptability of LLMs, domain-specific training remains paramount for their effective application. This work evaluates existing LLMs for recipe generation and proposes LLaVA-Chef, a novel model trained on a curated dataset of diverse recipe prompts in a multi-stage approach. First, we refine the mapping of visual food image embeddings to the language space. Second, we adapt LLaVA to the food domain by fine-tuning it on relevant recipe data. Third, we utilize diverse prompts to enhance the model's recipe comprehension. Finally, we improve the linguistic quality of generated recipes by penalizing the model with a custom loss function. LLaVA-Chef demonstrates impressive improvements over pretrained LLMs and prior works. A detailed qualitative analysis reveals that LLaVA-Chef generates more detailed recipes with precise ingredient mentions, compared to existing approaches.

The goal of the 22 nd International Workshop on Data Mining in Bioinformatics (BIOKDD 2023) is to encourage KDD researchers to solve the numerous problems and challenges in Bioinformatics using Data Mining technologies.Based on the organizers' expertise and communities, BIOKDD 2023 features the theme "Large-Scale Data-Driven Methods for Bioinformatics".This theme encourages the use of high-performance computing (HPC) to support the training of large machine learning models for problems in Bioinformatics and Computational Biology.The key goal is to accelerate the convergence between Data Mining and Bioinformatics communities to expedite discoveries in basic biology, medicine and healthcare.The goal of the 23 rd International Workshop on Data Mining in Bioinformatics (BIOKDD 2024) is to encourage KDD researchers to solve the numerous problems and challenges in Bioinformatics using Data Mining technologies.Based on the organizers' expertise and communities, BIOKDD 2024 features the theme "Advancing Bioinformatics with LLMs and GenAI".This theme encourages the use of large language models and generative artificial intelligence to solve problems in Bioinformatics and Computational Biology.The key goal is to accelerate the convergence between Data Mining and Bioinformatics communities to expedite discoveries in basic biology, medicine and healthcare.

In the financial sphere, there is a wealth of accumulated unstructured financial data, such as the textual disclosure documents that companies submit on a regular basis to regulatory agencies, such as the Securities and Exchange Commission. These documents are typically very long and tend to contain valuable soft information about a company’s performance that is not present in quantitative predictors. It is therefore of great interest to learn predictive models from these long textual documents, especially for forecasting numerical key performance indicators. In recent years, there has been great progress in natural language processing via pre-trained language models (LMs) learned from large corpora of textual data. This prompts the important question of whether they can be used effectively to produce representations for long documents, as well as how we can evaluate the effectiveness of representations produced by various LMs. Our work focuses on answering this critical question, namely, the evaluation of the efficacy of various LMs in extracting useful soft information from long textual documents for prediction tasks. In this article, we propose and implement a deep learning evaluation framework that utilizes a sequential chunking approach combined with an attention mechanism. We perform an extensive set of experiments on a collection of 10-K reports submitted annually by U.S. banks, and another dataset of reports submitted by U.S. companies, to investigate thoroughly the performance of different types of language models. Overall, our framework using LMs outperforms strong baseline methods for textual modeling as well as for numerical regression. Our work provides better insights into how utilizing pre-trained domain-specific and fine-tuned long-input LMs for representing long documents can improve the quality of representation of textual data and, therefore, help in improving predictive analyses.

The goal of the 22nd International Workshop on Data Mining in Bioinformatics (BIOKDD 2023) is to encourage KDD researchers to solve the numerous problems and challenges in Bioinformatics using Data Mining technologies. Based on the organizers' expertise and communities, BIOKDD 2023 features the theme ''Large-Scale Data-Driven Methods for Bioinformatics.'' This theme encourages the use of high-performance computing (HPC) to support the training of large machine learning models for problems in Bioinformatics and Computational Biology. The key goal is to accelerate the convergence between Data Mining and Bioinformatics communities to expedite discoveries in basic biology, medicine and healthcare.

Transformers use the dense self-attention mechanism which gives a lot of flexibility for long-range connectivity. Over multiple layers of a deep transformer, the number of possible connectivity patterns increases exponentially. However, very few of these contribute to the performance of the network, and even fewer are essential. We hypothesize that there are sparsely connected sub-networks within a transformer, called information pathways which can be trained independently. However, the dynamic (i.e., input-dependent) nature of these pathways makes it difficult to prune dense self-attention during training. But the overall distribution of these pathways is often predictable. We take advantage of this fact to propose Stochastically Subsampled self-Attention (SSA) - a general-purpose training strategy for transformers that can reduce both the memory and computational cost of self-attention by 4 to 8 times during training while also serving as a regularization method - improving generalization over dense training. We show that an ensemble of sub-models can be formed from the subsampled pathways within a network, which can achieve better performance than its densely attended counterpart. We perform experiments on a variety of NLP, computer vision and graph learning tasks in both generative and discriminative settings to provide empirical evidence for our claims and show the effectiveness of the proposed method.

The goal of the 21th International Workshop on Data Mining in Bioinformatics (BIOKDD 2022) is to encourage KDD researchers to solve the numerous problems and challenges in Bioinformatics using Data Mining technologies. Based on the organizers' expertise and communities, BIOKDD 2022 features 2 closely related themes "Biomedical Ontologies" and "Biological Data Visualization". These themes encourage the knowledge graph community and the data visualization community to work together to solve the many challenging problems in Bioinformatics with new perspectives. The key goal is to accelerate the convergence between Data Mining and Bioinformatics communities to expedite discoveries in basic biology, medicine and healthcare.

We propose an extension to the transformer neural network architecture for general-purpose graph learning by adding a dedicated pathway for pairwise structural information, called edge channels. The resultant framework - which we call Edge-augmented Graph Transformer (EGT) - can directly accept, process and output structural information of arbitrary form, which is important for effective learning on graph-structured data. Our model exclusively uses global self-attention as an aggregation mechanism rather than static localized convolutional aggregation. This allows for unconstrained long-range dynamic interactions between nodes. Moreover, the edge channels allow the structural information to evolve from layer to layer, and prediction tasks on edges/links can be performed directly from the output embeddings of these channels. We verify the performance of EGT in a wide range of graph-learning experiments on benchmark datasets, in which it outperforms Convolutional/Message-Passing Graph Neural Networks. EGT sets a new state-of-the-art for the quantum-chemical regression task on the OGB-LSC PCQM4Mv2 dataset containing 3.8 million molecular graphs. Our findings indicate that global self-attention based aggregation can serve as a flexible, adaptive and effective replacement of graph convolution for general-purpose graph learning. Therefore, convolutional local neighborhood aggregation is not an essential inductive bias.

Domain-specific keyword extraction is a vital task in the field of text mining. There are various research tasks, such as spam e-mail classification, abusive language detection, sentiment analysis, and emotion mining, where a set of domain-specific keywords (aka lexicon) is highly effective. Existing works for keyword extraction list all keywords rather than domain-specific keywords from a document corpus. Moreover, most of the existing approaches perform well on formal document corpuses but fail on noisy and informal user-generated content in online social media. In this article, we present a hybrid approach by jointly modeling the local and global contextual semantics of words, utilizing the strength of distributional word representation and contrasting-domain corpus for domain-specific keyword extraction. Starting with a seed set of a few domain-specific keywords, we model the text corpus as a weighted word-graph. In this graph, the initial weight of a node (word) represents its semantic association with the target domain calculated as a linear combination of three semantic association metrics, and the weight of an edge connecting a pair of nodes represents the co-occurrence count of the respective words. Thereafter, a modified PageRank method is applied to the word-graph to identify the most relevant words for expanding the initial set of domain-specific keywords. We evaluate our method over both formal and informal text corpuses (comprising six datasets), and show that it performs significantly better in comparison to state-of-the-art methods. Furthermore, we generalize our approach to handle the language-agnostic case, and show that it outperforms existing language-agnostic approaches.

The goal of the 20th International Workshop on Data Mining in Bioinformatics (BIOKDD 2021) is to encourage KDD researchers to tackle the numerous problems and challenges in Bioinformatics using Data Mining technologies. Based on the organizers' expertise and the BIOKDD communities, BIOKDD 2021 features the theme of "Artificial Intelligence in Medicine". This topic focuses on the use of machine learning and data mining techniques for the analysis of large amounts of heterogeneous, complex, biological and medical data, with a particular focus on deep learning methods that have seen rapid advance and wider adoption in Bioinformatics (e.g., DeepVariant, AlphaFold 2). We also particularly welcome COVID-19 related research. The key goal is to accelerate the convergence between Data Mining and Bioinformatics communities to expedite discoveries in basic biology, medicine and healthcare.

Food recommendation has become an important means to help guide users to adopt healthy dietary habits. Previous works on food recommendation either i) fail to consider users' explicit requirements, ii) ignore crucial health factors (e.g., allergies and nutrition needs), or iii) do not utilize the rich food knowledge for recommending healthy recipes. To address these limitations, we propose a novel problem formulation for food recommendation, modeling this task as constrained question answering over a large-scale food knowledge base/graph (KBQA). Besides the requirements from the user query, personalized requirements from the user's dietary preferences and health guidelines are handled in a unified way as additional constraints to the QA system. To validate this idea, we create a QA style dataset for personalized food recommendation based on a large-scale food knowledge graph and health guidelines. Furthermore, we propose a KBQA-based personalized food recommendation framework which is equipped with novel techniques for handling negations and numerical comparisons in the queries. Experimental results on the benchmark show that our approach significantly outperforms non-personalized counterparts (average 59.7% absolute improvement across various evaluation metrics), and is able to recommend more relevant and healthier recipes.

Recipe representation plays an important role in food computing for perception, recognition, recommendation and other applications. Learning pretrained recipe embeddings is a challenging task, as there is a lack of high quality annotated food datasets. In this paper, we provide a joint approach for learning effective pretrained recipe embeddings using both the ingredients and cooking instructions. We present RECIPTOR, a novel set transformer-based joint model to learn recipe representations, that preserves permutation-invariance for the ingredient set and uses a novel knowledge graph (KG) derived triplet sampling approach to optimize the learned embeddings so that related recipes are closer in the latent semantic space. The embeddings are further jointly optimized by combining similarity among cooking instructions with a KG based triplet loss. We experimentally show that RECIPTOR's recipe embeddings outperform state-of-the-art baselines on two newly designed downstream classification tasks by a wide margin.

The frequent usage of figurative language on online social networks, especially on Twitter, has the potential to mislead traditional sentiment analysis and recommender systems. Due to the extensive use of slangs, bashes, flames, and non-literal texts, tweets are a great source of figurative language, such as sarcasm, irony, metaphor, simile, hyperbole, humor, and satire. Starting with a brief introduction of figurative language and its various categories, this article presents an in-depth survey of the state-of-the-art techniques for computational detection of seven different figurative language categories, mainly on Twitter. For each figurative language category, we present details about the characterizing features, datasets, and state-of-the-art computational detection approaches. Finally, we discuss open challenges and future directions of research for each figurative language category.

Autoencoders have been successful in learning meaningful representations from image datasets. However, their performance on text datasets has not been widely studied. Traditional autoencoders tend to learn possibly trivial representations of text documents due to their confoundin properties such as high-dimensionality, sparsity and power-law word distributions. In this paper, we propose a novel k-competitive autoencoder, called KATE, for text documents. Due to the competition between the neurons in the hidden layer, each neuron becomes specialized in recognizing specific data patterns, and overall the model can learn meaningful representations of textual data. A comprehensive set of experiments show that KATE can learn better representations than traditional autoencoders including denoising, contractive, variational, and k-sparse autoencoders. Our model also outperforms deep generative models, probabilistic topic models, and even word representation models (e.g., Word2Vec) in terms of several downstream tasks such as document classification, regression, and retrieval.

Graph data mining is defined as searching in an input graph for all subgraphs that satisfy some property that makes them interesting to the user. Examples of graph data mining problems include frequent subgraph mining, counting motifs, and enumerating cliques. These problems differ from other graph processing problems such as PageRank or shortest path in that graph data mining requires searching through an exponential number of subgraphs. Most current parallel graph analytics systems do not provide good support for graph data mining. One notable exception is Arabesque, a system that was built specifically to support graph data mining. Arabesque provides a simple programming model to express graph data mining computations, and a highly scalable and efficient implementation of this model, scaling to billions of subgraphs on hundreds of cores. This demonstration will showcase the Arabesque system, focusing on the end-user experience and showing how Arabesque can be used to simply and efficiently solve practical graph data mining problems that would be difficult with other systems.

Frequent subgraph mining (FSM) has important applications in areas such as bioinformatics, social networks and others. In this paper, we present a highly scalable approach called ParGraph that can efficiently mine from a single graph in both distributed as well as shared-memory based systems. In a distributed environment, we can leverage the local memory of multiple compute nodes for storing a large number of intermediate states for enumerating patterns. To address the skewness in the pattern generation tree, our approach uses a novel hybrid load balancing scheme to efficiently distribute workload across both processes and threads. Our experiments demonstrate good speedups using message passing interface (MPI) and OpenMP threads.

The world is fast moving towards a data-driven society where data is the most valuable asset. Organizations need to perform very diverse analytic tasks using various data processing platforms. In doing so, they face many challenges; chiefly, platform dependence, poor interoperability, and poor performance when using multiple platforms. We present RHEEM, our vision for big data analytics over diverse data processing platforms. RHEEM provides a threelayer data processing and storage abstraction to achieve both platform independence and interoperability across multiple platforms. In this paper, we discuss our vision as well as present multiple research challenges that we need to address to achieve it. As a case in point, we present a data cleaning application built using some of the ideas of RHEEM. We show how it achieves platform independence and the performance benefits of following such an approach. 1. WHY TIED TO ONE SINGLE SYSTEM? Data analytic tasks may range from very simple to extremely complex pipelines, such as data extraction, transformation, and loading (ETL), online analytical processing (OLAP), graph processing, and machine learning (ML). Following the dictum “one size does not fit all” [23], academia and industry have embarked on an endless race to develop data processing platforms for supporting these different tasks, e.g., DBMSs and MapReduce-like systems. Semantic completeness, high performance, and scalability are key objectives of such platforms. While there have been major achievements in these objectives, users still face two main roadblocks. The first roadblock is that applications are tied to a single processing platform, making the migration of an application to new and more efficient platforms a difficult and costly task. Furthermore, complex analytic tasks usually require the combined use of different processing platforms. As a result, the common practice is to develop several specialized analytic applications on top of different platforms. This requires users to manually combine the results to draw a conclusion. In addition, users may need to re-implement existing applications on top of faster processing platforms when ∗Work done while at QCRI. c ©2016, Copyright is with the authors. Published in Proc. 19th International Conference on Extending Database Technology (EDBT), March 15-18, 2016 Bordeaux, France: ISBN 978-3-89318-070-7, on OpenProceedings.org. Distribution of this paper is permitted under the terms of the Creative Commons license CC-by-nc-nd 4.0 these become available. For example, Spark SQL [3] and MLlib [2] are the Spark counterparts of Hive [24] and Mahout [1]. The second roadblock is that datasets are often produced by different sources and hence they natively reside on different storage platforms. As a result, users often perform tedious, time-intensive, and costly data migration and integration tasks for further analysis. Let us illustrate these roadblocks with an Oil & Gas industry example [13]. A single oil company can produce more than 1.5TB of diverse data per day [6]. Such data may be structured or unstructured and come from heterogeneous sources, such as sensors, GPS devices, and other measuring instruments. For instance, during the exploration phase, data has to be acquired, integrated, and analyzed in order to predict if a reservoir would be profitable. Thousands of downhole sensors in exploratory wells produce real-time seismic data for monitoring resources and environmental conditions. Users integrate these data with the physical properties of the rocks to visualize volume and surface renderings. From these visualizations, geologists and geophysicists formulate hypotheses and verify them with ML methods, such as regression and classification. Training of the models is performed with historical drilling and production data, but oftentimes users have to go over unstructured data, such as notes exchanged by emails or text from drilling reports filed in a cabinet. Thus, an application supporting such a complex analytic pipeline has to access several sources for historical data (relational, but also text and semi-structured), remove the noise from the streaming data coming from the sensors, and run both traditional (such as SQL) and statistical analytics (such as ML algorithms) over different processing platforms. Similar examples can be drawn from many other domains such as healthcare: e.g., IBM reported that North York hospital needs to process 50 diverse datasets, which are on a dozen different internal systems [15]. These emerging applications clearly show the need for complex analytics coupled with a diversity of processing platforms, which raises two major research challenges. Data Processing Challenge. Users are faced with various choices on where to process their data, each choice with possibly orders of magnitude differences in terms of performance. However, users have to be intimate with the intricacies of the processing platform to achieve high efficiency and scalability. Moreover, once a decision is taken, users may end up being tied up to a particular platform. As a result, migrating the data analytics stack to a more efficient processing platform often becomes a nightmare. Thus, there is a need to build a system that offers data processing platform independence. Furthermore, complex analytic applications require executing tasks over different processing platforms to achieve high performance. For example, one may aggregate large datasets with traditional queries on top of a relational database such as PostgreSQL, but ML tasks might be much faster if executed on Spark [28]. HowVisionary Paper Series ISSN: 2367-2005 479 10.5441/002/edbt.2016.45 ever, this requires a considerable amount of manual work in selecting the best processing platforms, optimizing tasks for the chosen platforms, and coordinating task execution. Thus, this also calls for multi-platform task execution. Data Storage Challenge. Data processing platforms are typically tightly coupled with a specific storage solution. Moving data from a certain storage (e.g., a relational DB) to a more suitable processing platform for the actual task (e.g., Spark on HDFS) requires shuffling data between different systems. Such shuffling may end up dominating the execution time. Moreover, different departments in the same organization may go for different storage engines due to legacy as well as performance reasons. Dealing with such heterogeneity calls for data storage independence. To tackle these two challenges, we envision a system, called RHEEM1, that provides both platform independence and interoperability (Section 2). In the following, we first discuss our vision for the data processing abstraction (Section 3), which is fully based on user-defined functions (UDFs) to provide adaptability as well as extensibility. This processing abstraction allows both users to focus only on the logic of their data analytic tasks and applications to be independent from the data processing platforms. We then discuss how to divide a complex analytic task into smaller subtasks to exploit the availability of different processing platforms (Section 4). As a result, RHEEM can run simultaneously a single data analytic task over multiple processing platforms to boost performance. Next, we present our first attempt to build an instance application based on some of the ideas of RHEEM and the resulting benefits (Section 5). We then show how we push down the processing abstraction idea to the storage layer (Section 6). This storage abstraction allows both users to focus on their storage needs and the processing platforms to be independent from the storage engines. Some initial efforts are also going into the direction of providing data processing platform independence [11,12,21] (Section 7). However, our vision goes beyond the data processing. We not only envision a data processing abstraction but also a data storage abstraction, allowing us to consider data movement costs during task optimization. We give a research agenda highlighting the challenges that need to be tackled to build RHEEM in Section 8.

Many emerging applications, from domains such as healthcare and oil & gas, require several data processing systems for complex analytics. This demo paper showcases system, a framework that provides multi-platform task execution for such applications. It features a three-layer data processing abstraction and a new query optimization approach for multi-platform settings. We will demonstrate the strengths of system by using real-world scenarios from three different applications, namely, machine learning, data cleaning, and data fusion.

A key problem in many graph-based applications is the need to know, given a directed graph G and two vertices u,v ∈ G, whether there is a path between u and v, i.e., if u reaches v. This problem is particularly challenging in the case of very large real-world graphs. A common approach is the preprocessing of the graphs, in order to produce an efficient index structure, which allows fast access to the reachability information of the vertices. However, the majority of existing methods can not handle very large graphs. We propose, in this paper, a novel indexing method called FELINE (Fast rEfined onLINE search), which is inspired by Dominance Graph Drawing. FELINE creates an index from the graph representation in a two-dimensional plane, which provides reachability information in constant time for a significant portion of queries. Experiments demonstrate the efficiency of FELINE compared to state-of-the-art approaches.

Many real-world graphs have complex labels on the nodes and edges. Mining only exact patterns yields limited insights, since it may be hard to find exact matches. However, in many domains it is relatively easy to define a cost (or distance) between different labels. Using this information, it becomes possible to mine a much richer set of approximate subgraph patterns, which preserve the topology but allow bounded label mismatches. We present novel and scalable methods to efficiently solve the approximate isomorphism problem. We show that approximate mining yields interesting patterns in several real-world graphs ranging from IT and protein interaction networks to protein structures.

We tackle the challenging problem of mining the simplest Boolean patterns from categorical datasets. Instead of complete enumeration, which is typically infeasible for this class of patterns, we develop effective sampling methods to extract a representative subset of the minimal Boolean patterns (in disjunctive normal form - DNF). We make both theoretical and practical contributions, which allow us to prune the search space based on provable properties. Our approach can provide a near-uniform sample of the minimal DNF patterns. We also show that the mined minimal DNF patterns are very effective when used as features for classification.

In this work, we study the correlation between attribute sets and the occurrence of dense subgraphs in large attributed graphs, a task we call structural correlation pattern mining. A structural correlation pattern is a dense subgraph induced by a particular attribute set. Existing methods are not able to extract relevant knowledge regarding how vertex attributes interact with dense subgraphs. Structural correlation pattern mining combines aspects of frequent itemset and quasi-clique mining problems. We propose statistical significance measures that compare the structural correlation of attribute sets against their expected values using null models. Moreover, we evaluate the interestingness of structural correlation patterns in terms of size and density. An efficient algorithm that combines search and pruning strategies in the identification of the most relevant structural correlation patterns is presented. We apply our method for the analysis of three real-world attributed graphs: a collaboration, a music, and a citation network, verifying that it provides valuable knowledge in a feasible time.

A configuration management database (CMDB) can be considered to be a large graph representing the IT infrastructure entities and their inter-relationships. Mining such graphs is challenging because they are large, complex, and multi-attributed, and have many repeated labels. These characteristics pose challenges for graph mining algorithms, due to the increased cost of sub graph isomorphism (for support counting), and graph isomorphism (for eliminating duplicate patterns). The notion of pattern frequency or support is also more challenging in a single graph, since it has to be defined in terms of the number of its (potentially, exponentially many) embeddings. We present CMDB-Miner, a novel two-step method for mining infrastructure patterns from CMDB graphs. It first samples the set of maximal frequent patterns, and then clusters them to extract the representative infrastructure patterns. We demonstrate the effectiveness of CMDB-Miner on real-world CMDB graphs.

A wide variety of clustering algorithms exist that cater to applications based on certain special characteristics of the data. Our focus is on methods that capture arbitrary shaped clusters in data, the so called spatial clustering algorithms. With the growing size of spatial datasets from diverse sources, the need for scalable algorithms is paramount. We propose a shape-based clustering algorithm, ABACUS, that scales to large datasets. ABACUS is based on the idea of identifying the intrinsic structure for each cluster, which we also refer to as the backbone of that cluster. The backbone comprises of a much smaller set of points, thus giving this method the desired ability to scale to larger datasets. ABACUS operates in two stages. In the first stage, we identify the backbone of each cluster via an iterative process made up of globbing (or point merging) and point movement operations. The backbone enables easy identification of the true clusters in a subsequent stage. Experiments on a range of real (images from geospatial satellites, etc.) and synthetic datasets demonstrate the efficiency and effectiveness of our approach. In particular, ABACUS is over an order of magnitude faster than existing shape-based clustering methods, yet it provides a comparable or better clustering quality.

When multiple data sources are available for clustering, an a priori data integration process is usually required. This process may be costly and may not lead to good clusterings, since important information is likely to be discarded. In this paper we propose constrained clustering as a strategy for integrating data sources without losing any information. It basically consists of adding the complementary data sources as constraints that the algorithm must satisfy. As a concrete application of our approach, we focus on the problem of enzyme function prediction, which is a hard task usually performed by intensive experimental work. We use constrained clustering as a means of integrating information from diverse sources as constraints, and analyze how this additional information impacts clustering quality in an enzyme clustering application scenario. Our results show that constraints generally improve the clustering quality when compared to an unconstrained clustering algorithm.

The Semantic Web community, until now, has used traditional database systems for the storage and querying of RDF data. The SPARQL query language also closely follows SQL syntax. As a natural consequence, most of the SPARQL query processing techniques are based on database query processing and optimization techniques. For SPARQL join query optimization, previous works like RDF-3X and Hexastore have proposed to use 6-way indexes on the RDF data. Although these indexes speed up merge-joins by orders of magnitude, for complex join queries generating large intermediate join results, the scalability of the query processor still remains a challenge.

Given a large directed graph, rapidly answering reachability queries between source and target nodes is an important problem. Existing methods for reachability trade-off indexing time and space versus query time performance. However, the biggest limitation of existing methods is that they simply do not scale to very large real-world graphs. We present a very simple, but scalable reachability index, called GRAIL, that is based on the idea of randomized interval labeling, and that can effectively handle very large graphs. Based on an extensive set of experiments, we show that while more sophisticated methods work better on small graphs, GRAIL is the only index that can scale to millions of nodes and edges. GRAIL has linear indexing time and space, and the query time ranges from constant time to being linear in the graph order and size.

We present VOGUE, a novel, variable order hidden Markov model with state durations, that combines two separate techniques for modeling complex patterns in sequential data: pattern mining and data modeling. VOGUE relies on a variable gap sequence mining method to extract frequent patterns with different lengths and gaps between elements. It then uses these mined sequences to build a variable order hidden Markov model (HMM), that explicitly models the gaps. The gaps implicitly model the order of the HMM, and they explicitly model the duration of each state. We apply VOGUE to a variety of real sequence data taken from domains such as protein sequence classification, Web usage logs, intrusion detection, and spelling correction. We show that VOGUE has superior classification accuracy compared to regular HMMs, higher-order HMMs, and even special purpose HMMs like HMMER, which is a state-of-the-art method for protein classification. The VOGUE implementation and the datasets used in this article are available as open-source. 1

Recent research in frequent pattern mining (FPM) has shifted from obtaining the complete set of frequent patterns to generating only a representative (summary) subset of frequent patterns. Most of the existing approaches to this problem adopt a two-step solution; in the first step, they obtain all the frequent patterns, and in the second step, some form of clustering is used to obtain the summary pattern set. However, the two-step method is inefficient and sometimes infeasible since the first step itself may fail to finish in a reasonable amount of time. In this paper, we propose an alternative approach to mining frequent pattern representatives based on a uniform sampling of the output space. Our new algorithm, Musk, obtains representative patterns by sampling uniformly from the pool of all frequent maximal patterns; uniformity is achieved by a variant of Markov Chain Monte Carlo (MCMC) algorithm. Musk simulates a random walk on the frequent pattern partial order graph with a prescribed transition probability matrix, whose values are computed locally during the simulation. In the stationary distribution of the random walk, all maximal frequent pattern nodes in the partial order graph are sampled uniformly. Experiments on various kind of graph and itemset databases validate the effectiveness of our approach.

The classification performance of an associative classifier is strongly dependent on the statistic measure or metric that is used to quantify the strength of the association between features and classes (i.e., confidence, correlation etc.). Previous studies have shown that classifiers produced by different metrics may provide conflicting predictions, and that the best metric to use is data-dependent and rarely known while designing the classifier. This uncertainty concerning the optimal match between metrics and problems is a dilemma, and prevents associative classifiers to achieve their maximal performance. This dilemma is the focus of this paper. A possible solution to this dilemma is to learn the competence, expertise, or assertiveness of metrics. The basic idea is that each metric has a specific sub-domain for which it is most competent (i.e., it consistently produces more accurate classifiers than the ones produced by other metrics). Particularly, we investigate stacking-based meta-learning methods, which use the training data to find the domain of competence of each metric. The meta-classifier describes the domains of competence (or areas of expertise) of each metric, enabling a more sensible use of these metrics so that competence-conscious classifiers can be produced (i.e., a metric is only used to produce classifiers for test instances that belong to its domain of competence). We conducted a systematic evaluation, using different datasets and evaluation measures, of classifiers produced by different metrics. The result is that, while no metric is always superior than all others, the selection of appropriate metrics according to their competence/expertise (i.e., competence-conscious associative classifiers) seems very effective, showing gains that range from 7% to 26% when compared to the baselines (SVMs and an existing ensemble method).

Recent interest in graph pattern mining has shifted from finding all frequent subgraphs to obtaining a small subset of frequent subgraphs that are representative, discriminative or significant. The main motivation behind that is to cope with the scalability problem that the graph mining algorithms suffer when mining databases of large graphs. Another motivation is to obtain a succinct output set that is informative and useful. In the same spirit, researchers also proposed sampling based algorithms that sample the output space of the frequent patterns to obtain representative subgraphs. In this work, we propose a generic sampling framework that is based on Metropolis-Hastings algorithm to sample the output space of frequent subgraphs. Our experiments on various sampling strategies show the versatility, utility and efficiency of the proposed sampling approach.

Clustering is one of the fundamental data mining tasks. Many different clustering paradigms have been developed over the years, which include partitional, hierarchical, mixture model based, density-based, spectral, subspace, and so on. The focus of this paper is on full-dimensional, arbitrary shaped clusters. Existing methods for this problem suffer either in terms of the memory or time complexity (quadratic or even cubic). This shortcoming has restricted these algorithms to datasets of moderate sizes. In this paper we propose SPARCL, a simple and scalable algorithm for finding clusters with arbitrary shapes and sizes, and it has linear space and time complexity. SPARCL consists of two stages - the first stage runs a carefully initialized version of the K-means algorithm to generate many small seed clusters. The second stage iteratively merges the generated clusters to obtain the final shape-based clusters. Experiments were conducted on a variety of datasets to highlight the effectiveness, efficiency, and scalability of our approach. On the large datasets SPARCL is an order of magnitude faster than the best existing approaches.

No abstract available.

Sequence mining is one of the fundamental data mining tasks. In this paper we present a novel approach called Prism, for mining frequent sequences. Prism utilizes a vertical approach for enumeration and support counting, based on the novel notion o/prime block encoding, which in turn is based on prime factorization theory. Via an extensive evaluation on both synthetic and real datasets, we show that Prism outperforms popular sequence mining methods like SPADE [10], PrefixSpan [6] and SPAM [2], by an order of magnitude or more.

In this paper, we introduce the concept of alpha-orthogonal patterns to mine a representative set of graph patterns. Intuitively, two graph patterns are alpha-orthogonal if their similarity is bounded above by alpha. Each alpha-orthogonal pattern is also a representative for those patterns that are at least beta similar to it. Given user defined alpha, beta isin [0,1], the goal is to mine an alpha-orthogonal, beta-representative set that minimizes the set of unrepresented patterns. We present ORIGAMI, an effective algorithm for mining the set of representative orthogonal patterns. ORIGAMI first uses a randomized algorithm to randomly traverse the pattern space, seeking previously unexplored regions, to return a set of maximal patterns. ORIGAMI then extracts an alpha-orthogonal, beta-representative set from the mined maximal patterns. We show the effectiveness of our algorithm on a number of real and synthetic datasets. In particular, we show that our method is able to extract high quality patterns even in cases where existing enumerative graph mining methods fail to do so.

XML has become a popular method of data representation both on the web and in databases in recent years. One of the reasons for the popularity of XML has been its ability to encode structural information about data records. However, this structural characteristic of data sets also makes it a challenging problem for a variety of data mining problems. One such problem is that of clustering, in which the structural aspects of the data result in a high implicit dimensionality of the data representation. As a result, it becomes more difficult to cluster the data in a meaningful way. In this paper, we propose an effective clustering algorithm for XML data which uses substructures of the documents in order to gain insights about the important underlying structures. We propose new ways of using multiple sub-structuralinformation in XML documents to evaluate the quality of intermediate cluster solutions, and guide the algorithms to a final solution which reflects the true structural behavior in individual partitions. We test the algorithm on a variety of real and synthetic data sets.

With the exponential growth of biological sequence databases, it has become critical to develop effective techniques for storing, querying, and analyzing these massive data. Suffix trees are widely used to solve many sequence-based problems, and they can be built in linear time and space, provided the resulting tree fits in main-memory. To index larger sequences, several external suffix tree algorithms have been proposed in recent years. However, they suffer from several problems such as susceptibility to data skew, non-scalability to genome-scale sequences, and non-existence of suffix links, which are crucial in various suffix tree based algorithms. In this paper, we target DNA sequences and propose a novel disk-based suffix tree algorithm called TRELLIS, which effectively scales up to genome-scale sequences. Specifically, it can index the entire human genome using 2GB of memory, in about 4 hours and can recover all its suffix links within 2 hours. TRELLIS was compared to various state-of-the-art persistent disk-based suffix tree construction algorithms, and was shown to outperform the best previous methods, both in terms of indexing time and querying time.

We present a novel approach for classifying documents that combines different pieces of evidence (e.g., textual features of documents, links, and citations) transparently, through a data mining technique which generates rules associating these pieces of evidence to predefined classes. These rules can contain any number and mixture of the available evidence and are associated with several quality criteria which can be used in conjunction to choose the "best" rule to be applied at classification time. Our method is able to perform evidence enhancement by link forwarding/backwarding (i.e., navigating among documents related through citation), so that new pieces of link-based evidence are derived when necessary. Furthermore, instead of inducing a single model (or rule set) that is good on average for all predictions, the proposed approach employs a lazy method which delays the inductive process until a document is given for classification, therefore taking advantage of better qualitative evidence coming from the document. We conducted a systematic evaluation of the proposed approach using documents from the ACM Digital Library and from a Brazilian Web directory. Our approach was able to outperform in both collections all classifiers based on the best available evidence in isolation as well as state-of-the-art multi-evidence classifiers. We also evaluated our approach using the standard WebKB collection, where our approach showed gains of 1% in accuracy, being 25 times faster. Further, our approach is extremely efficient in terms of computational performance, showing gains of more than one order of magnitude when compared against other multi-evidence classifiers.

Decision tree classifiers perform a greedy search for rules by heuristically selecting the most promising features. Such greedy (local) search may discard important rules. Associative classifiers, on the other hand, perform a global search for rules satisfying some quality constraints (i.e., minimum support). This global search, however, may generate a large number of rules. Further, many of these rules may be useless during classification, and worst, important rules may never be mined. Lazy (non-eager) associative classification overcomes this problem by focusing on the features of the given test instance, increasing the chance of generating more rules that are useful for classifying the test instance. In this paper we assess the performance of lazy associative classification. First we demonstrate that an associative classifier performs no worse than the corresponding decision tree classifier. Also we demonstrate that lazy classifiers outperform the corresponding eager ones. Our claims are empirically confirmed by an extensive set of experimental results. We show that our proposed lazy associative classifier is responsible for an error rate reduction of approximately 10 % when compared against its eager counterpart, and for a reduction of 20 % when compared against a decision tree classifier. A simple caching mechanism makes lazy associative classification fast, and thus improvements in the execution time are also observed. 1

International audience

We introduce a novel framework, called BLOSOM, for mining (frequent) boolean expressions over binary-valued datasets. We organize the space of boolean expressions into four categories: pure conjunctions, pure disjunctions, conjunction of disjunctions, and disjunction of conjunctions. We focus on mining the simplest expressions the minimal generators for each class. We also propose a closure operator for each class that yields closed boolean expressions. BLOSOM efficiently mines frequent boolean expressions by utilizing a number of methodical pruning techniques. Experiments showcase the behavior of BLOSOM, and an application study on a real dataset is also given.

We present a novel algorithm called CLICKS, that finds clusters in categorical datasets based on a search for k-partite maximal cliques. Unlike previous methods, CLICKS mines subspace clusters. It uses a selective vertical method to guarantee complete search. CLICKS outperforms previous approaches by over an order of magnitude and scales better than any of the existing method for high-dimensional datasets. We demonstrate this improvement in an excerpt from our comprehensive performance studies.

The resource requirements of frequent pattern mining algorithms depend mainly on the length distribution of the mined patterns in the database. Synthetic databases, which are used to benchmark performance of algorithms, tend to have distributions far different from those observed in real datasets. In this paper we focus on the problem of synthetic database generation and propose algorithms to effectively embed within the database, any given set of maximal pattern collections, and make the following contributions: 1. A database generation technique is presented which takes k maximal itemset collections as input, and constructs a database which produces these maximal collections as output, when mined at k levels of support. To analyze the efficiency of the procedure, upper bounds are provided on the number of transactions output in the generated database; 2. A compression method is used and extended to reduce the size of the output database. An optimization to the generation procedure is provided which could potentially reduce the number of transactions generated; 3. Preliminary experimental results are presented to demonstrate the feasibility of using the generation technique.

We present a novel algorithm called CLICKS, that finds clusters in categorical datasets based on a search for k-partite maximal cliques. Unlike previous methods, CLICKS mines subspace clusters. It uses a selective vertical method to guarantee complete search. CLICKS outperforms previous approaches by over an order of magnitude and scales better than any of the existing method for high-dimensional datasets. These results are demonstrated in a comprehensive performance study on real and synthetic datasets.

Redescription mining is a newly introduced data mining problem that seeks to find subsets of data that afford multiple definitions. It can be viewed as a generalization of association rule mining, from finding implications to equivalences; as a form of conceptual clustering, where the goal is to identify clusters that afford dual characterizations; and as a form of constructive induction, to build features based on given descriptors that mutually reinforce each other. In this paper, we present the use of redescription mining as an important tool to reason about a collection of sets, especially their overlaps, similarities, and differences. We outline algorithms to mine all minimal (non-redundant) redescriptions underlying a dataset using notions of minimal generators of closed itemsets. We also show the use of these algorithms in an interactive context, supporting constraint-based exploration and querying. Specifically, we showcase a bioinformatics application that empowers the biologist to define a vocabulary of sets underlying a domain of genes and to reason about these sets, yielding significant biological insight.

In this paper we introduce a novel algorithm called TRICLUSTER, for mining coherent clusters in three-dimensional (3D) gene expression datasets. TRICLUSTER can mine arbitrarily positioned and overlapping clusters, and depending on different parameter values, it can mine different types of clusters, including those with constant or similar values along each dimension, as well as scaling and shifting expression patterns. TRICLUSTER relies on graph-based approach to mine all valid clusters. For each time slice, i.e., a gene×sample matrix, it constructs the range multigraph, a compact representation of all similar value ranges between any two sample columns. It then searches for constrained maximal cliques in this multigraph to yield the set of bi-clusters for this time slice. Then TRICLUSTER constructs another graph using the biclusters (as vertices) from each time slice; mining cliques from this graph yields the final set of triclusters. Optionally, TRICLUSTER merges/deletes some clusters having large overlaps. We present a useful set of metrics to evaluate the clustering quality, and we show that TRICLUSTER can find significant triclusters in the real microarray datasets.

Mining frequent trees is very useful in domains like bioinformatics, Web mining, mining semistructured data, etc. We formulate the problem of mining (embedded) subtrees in a forest of rooted, labeled, and ordered trees. We present TREEMINER, a novel algorithm to discover all frequent subtrees in a forest, using a new data structure called scope-list. We contrast TREEMINER with a pattern matching tree mining algorithm (PATTERNMATCHER), and we also compare it with TREEMINERD, which counts only distinct occurrences of a pattern. We conduct detailed experiments to test the performance and scalability of these methods. We also use tree mining to analyze RNA structure and phylogenetics data sets from bioinformatics domain.

The set of frequent closed itemsets uniquely determines the exact frequency of all itemsets, yet it can be orders of magnitude smaller than the set of all frequent itemsets. In this paper, we present CHARM, an efficient algorithm for mining all frequent closed itemsets. It enumerates closed sets using a dual itemset-tidset search tree, using an efficient hybrid search that skips many levels. It also uses a technique called diffsets to reduce the memory footprint of intermediate computations. Finally, it uses a fast hash-based approach to remove any "nonclosed" sets found during computation. We also present CHARM-L, an algorithm that outputs the closed itemset lattice, which is very useful for rule generation and visualization. An extensive experimental evaluation on a number of real and synthetic databases shows that CHARM is a state-of-the-art algorithm that outperforms previous methods. Further, CHARM-L explicitly generates the frequent closed itemset lattice.

High-dimensional data pose challenges to traditional clustering algorithms due to their inherent sparsity and data tend to cluster in different and possibly overlapping subspaces of the entire feature space. Finding such subspaces is called subspace mining. We present SCHISM, a new algorithm for mining interesting subspaces, using the notions of support and Chernoff-Hoeffding bounds. We use a vertical representation of the dataset, and use a depth-first search with backtracking to find maximal interesting subspaces. We test our algorithm on a number of high-dimensional synthetic and real datasets to test its effectiveness.

The growth of bioinformatics has resulted in datasets with new characteristics. These datasets typically contain a large number of columns and a small number of rows. For example, many gene expression datasets may contain 10,000-100,000 columns but only 100-1000 rows.Such datasets pose a great challenge for existing (closed) frequent pattern discovery algorithms, since they have an exponential dependence on the average row length. In this paper, we describe a new algorithm called CARPENTER that is specially designed to handle datasets having a large number of attributes and relatively small number of rows. Several experiments on real bioinformatics datasets show that CARPENTER is orders of magnitude better than previous closed pattern mining algorithms like CLOSET and CHARM.

In most computer systems, page fault rate is currently minimized by generic page replacement algorithms which try to model the temporal locality inherent in programs. In this paper, we propose two algorithms, one greedy and the other stochastic, designed for program specific code restructuring as a means of increasing spatial locality within a program. Both algorithms e#ectively decrease average working set size and hence the page fault rate. Our methods are more e#ective than traditional approaches due to use of domain information. We illustrate the e#cacy of our algorithms on actual data mining algorithms.

XML documents have recently become ubiquitous because of their varied applicability in a number of applications. Classification is an important problem in the data mining domain, but current classification methods for XML documents use IR-based methods in which each document is treated as a bag of words. Such techniques ignore a significant amount of information hidden inside the documents. In this paper we discuss the problem of rule based classification of XML data by using frequent discriminatory substructures within XML documents. Such a technique is more capable of finding the classification characteristics of documents. In addition, the technique can also be extended to cost sensitive classification. We show the effectiveness of the method with respect to other classifiers. We note that the methodology discussed in this paper is applicable to any kind of semi-structured data.

A number of vertical mining algorithms have been proposed recently for association mining, which have shown to be very effective and usually outperform horizontal approaches. The main advantage of the vertical format is support for fast frequency counting via intersection operations on transaction ids (tids) and automatic pruning of irrelevant data. The main problem with these approaches is when intermediate results of vertical tid lists become too large for memory, thus affecting the algorithm scalability. In this paper we present a novel vertical data representation called Diffset, that only keeps track of differences in the tids of a candidate pattern from its generating frequent patterns. We show that diffsets drastically cut down the size of memory required to store intermediate results. We show how diffsets, when incorporated into previous vertical mining methods, increase the performance significantly. We also present a new algorithm, using diffsets, for mining maximal patterns. Experimental comparisons, on both dense and sparse databases, show that diffsets deliver order of magnitude performance improvements over the best previous methods. 1

Computing frequent itemsets and maximally frequent item-sets in a database are classic problems in data mining. The resource requirements of all extant algorithms for both problems depend on the distribution of frequent patterns, a topic that has not been formally investigated. In this paper, we study properties of length distributions of frequent and maximal frequent itemset collections and provide novel solutions for computing tight lower bounds for feasible distributions. We show how these bounding distributions can help in generating realistic synthetic datasets, which can be used for algorithm benchmarking.

Security of computer systems is essential to their acceptance and utility. Computer security analysts use intrusion detection systems to assist them in maintaining computer system security. This paper deals with the problem of differentiating between masqueraders and the true user of a computer terminal. Prior efficient solutions are less suited to real time application, often requiring all training data to be labeled, and do not inherently provide an intuitive idea of what the data model means. Our system, called ADMIT, relaxes these constraints, by creating user profiles using semi-incremental techniques. It is a real-time intrusion detection system with host-based data collection and processing. Our method also suggests ideas for dealing with concept drift and affords a detection rate as high as 80.3% and a false positive rate as low as 15.3%.

A$ 12'34,&$\t20&$\t-&$5/67 1 ##(89(:2(8;#< N(< #LM, 28950-47690 \t-&$5/67 1 ##(89(:2(8;#<%=%'?> V=WKXXYAZ[QXRW\\2(]%0";('"\t.%8 C&^%#1%0 O('""; (89(:2(8;#<%=%' *J(F (]%0";('"\t.%8 C&^%#1%0 O('""; (89(:2(8;#<%=%'?> 29099 @AO1%. 9(87OVMWKXRXYAZ[QXRWJ3-C)(D(8 J&)(819C\tM O&$R/ \tb('"\t.%8 5C&cGadPe:f;f;XWK[PYge:fKh;iQXRWQL >j@H1%.#1 )#9(8"# kDK &$:*%<74C-DK Yge:fKh;iQXRWQL 19490-42850 \tM O&$R/ 2 &$ C%:#>@H$m;#gC;(0$VMWKXRXYAZ[QXRWH%.DK :#>@H$m;#gC; $&)(819C\t<(8DDR %('19C$,T( XRWH%.DK 22620-40920 \tM O&$ 1('"DbDR %.DK >q@A)('"%rD :B(8s('DD"1(8%.b%8 / 29000 &$\tB%$(8;(8"Rt ('"?34,u"%.\t'F :B(8s('DD"1(8% 1.

1 Introduction The field of knowledge discovery and data mining (KDD), spurred by advances in data collection technology, is concerned with the process of deriving interesting and useful patterns from large datasets. The KDD process is computational and data-intensive and is inherently interactive and iterative in nature. In fact, interactivity is often the key to facilitating effective data understanding and knowledge discovery. In such an environment, response time is crucial because lengthy time delay between responses of consecutive user requests can disturb the flow of human perception and formation of insight. The task of guaranteeing quick response times is more complicated in dynamic datasets, where there is a constant influx of data. Changes to the data can invalidate existing patterns or introduce new. Simply re-executing algorithms from scratch when a database is updated can result in an explosion in the computational and I/O resources required. What is needed is a way to process the data incrementally and update the information that is gleaned while being cognizant of the interactive requirements of the process. In this paper we present such an approach for a key data mining task: association rule mining.

The set of frequent closed itemsets uniquely determines the exact frequency of all itemsets, yet it can be orders of magnitude smaller than the set of all frequent itemsets. In this paper we present CHARM, an efficient algorithm for mining all frequent closed itemsets. It enumerates closed sets using a dual itemset-tidset search tree, using an efficient hybrid search that skips many levels. It also uses a technique called diffsets to reduce the memory footprint of intermediate computations. Finally it uses a fast hash-based approach to remove any “non-closed” sets found during computation. An extensive experimental evaluation on a number of real and synthetic databases shows that CHARM significantly outperforms previous methods. It is also linearly scalable in the number of transactions.

We present GenMax, a backtracking search based algorithm for mining maximal frequent itemsets. GenMax uses a number of optimizations to prune the search space. It uses a novel technique called progressive focusing to perform maximality checking, and diffset propagation to perform fast frequency computation. Systematic experimental comparison with previous work indicates that different methods have varying strengths and weaknesses based on dataset characteristics. We found GenMax to be a highly efficient method to mine the exact set of maximal patterns.

We present cSPADE, an efficient algorithm for mining frequent sequences considering a variety of syntactic constraints.These take the form of length or width limitations on the sequences, minimum or maximum gap constraints on consecutive sequence elements, applying a time window on allowable sequences, incorporating item constraints, and finding sequences predictive of one or more classes, even rare ones.Our method is efficient and scalable.Experiments on a number of synthetic and real databases show the utility and performance of considering such constraints on the set of mined sequences. INTRODUCTIONThis paper focuses on sequence data in which each example is represented as a sequence of "events", where each event might be described by a set of predicates, i.e., we are dealing with categorical sequential domains.Examples of sequence data include text, DNA sequences, web usage data, multi-player games, plan execution traces, and so on.The sequence mining task is to discover a sequence of attributes, shared across time among a large number of objects in a given database.For example, consider a web access database at a popular site, where an object is a web user and an attribute is a web page.The discovered patterns are the sequences of most frequently accessed pages at that site.This kind of information can be used to restructure the web-site, or to dynamically insert relevant links in web pages based on user access patterns.There are many other domains where sequence mining has been applied, which include discovering customer buying patterns in retail stores, identifying plan failures [12], finding network alarms [3], and so on.The task of discovering all frequent sequences in large databases is quite challenging.The search space is extremely large.For example, with m attributes there are O(m k ) potentially frequent sequences of length at most k.Many techniques have been proposed to mine temporal databases for the frequently occurring sequences.However, an unconstrained search can produce millions of rules or may even be intractable in some domains.Furthermore, in many do-

The traditional association rule mining framework produces many redundant rules. The extent of redundancy is a lot larger than previously suspected. We present a new framework for associations based on the concept of closed frequent itemsets. The number of non-redundant rules produced by the new approach is exponentially (in the length of the longest frequent itemset) smaller than the rule set from the traditional approach. Experiments using several "hard" as well as "easy" real and synthetic databases confirm the utility of our framework in terms of reduction in the number of rules presented to the user, and in terms of time.

Association rule discovery has emerged as an important problem in knowledge discovery and data mining. The association mining task consists of identifying the frequent itemsets, and then forming conditional implication rules among them. We present efficient algorithms for the discovery of frequent itemsets which forms the compute intensive phase of the task. The algorithms utilize the structural properties of frequent itemsets to facilitate fast discovery. The items are organized into a subset lattice search space, which is decomposed into small independent chunks or sublattices, which can be solved in memory. Efficient lattice traversal techniques are presented which quickly identify all the long frequent itemsets and their subsets if required. We also present the effect of using different database layout schemes combined with the proposed decomposition and traversal techniques. We experimentally compare the new algorithms against the previous approaches, obtaining improvements of more than an order of magnitude for our test databases.

The discovery of frequent sequences in temporal databases is an important data mining problem. Most current work assumes that the database is static, and a database update requires rediscovering all the patterns by scanning the entire old and new database. In this paper, we propose novel techniques for maintaining sequences in the presence of a) database updates, and b) user interaction (e.g. modifying mining parameters). This is a very challenging task, since such updates can invalidate existing sequences or introduce new ones. In both the above scenarios, we avoid re-executing the algorithm on the entire dataset, thereby reducing execution time. Experimental results confirm that our approach results in execution time improvements of up to several orders of magnitude in practice.

Presents parallel algorithms for building decision-tree classifiers on shared-memory multiprocessor (SMP) systems. The proposed algorithms span the gamut of data and task parallelism. The data parallelism is based on attribute scheduling among processors. This basic scheme is extended with task pipelining and dynamic load balancing to yield faster implementations. The task-parallel approach uses dynamic subtree partitioning among processors. Our performance evaluation shows that the construction of a decision-tree classifier can be effectively parallelized on an SMP machine with good speedup.

Classification algorithms are difficult to apply to sequential examples because there is a vast number of potentially useful features for describing each example.Past work on feature selection has focused on searching the space of all subsets of features, which is intractable for large feature sets.We adapt sequence mining techniques to aEi as a preprocessor to select features for standard classification algorithms such as Naive Bayes and Winnow.Our experiments on three different datasets show that the features produced by our algorithm improve classification accuracy by lo-50%,

In this paper we present SPADE, a new algorithm for fast discovery of Sequential Patterns. The existing solutions to this problem make repeated database scans, and use complex hash structures which have poor locality. SPADE utilizes combinatorial properties to decompose the original problem into smaller sub-problems, that can be independently solved in main-memory using efficient lattice search techniques, and using simple join operations. All sequences are discovered in only three database scans. Experiments show that SPADE outperforms the best previous algorithm by a factor of two, and by an order of magnitude with some pre-processed data. It also has linear scalability with respect to the number of customers, and a number of other database parameters.