Twitter is currently one of the most influential microblogging services on which users interact with messages. It is imperative to grasp the big picture of Twitter through analyzing its huge stream data. In this study, we develop a two-stage clustering method that automatically discovers coarse-grained topics from Twitter data. In the first stage, we use graph clustering to extract micro-clusters from the word co-occurrence graph. All the tweets in a micro-cluster share a fine-grained topic. We then obtain the time series of each micro-cluster by counting the number of tweets posted in a time window. In the second stage, we use time series clustering to identify the clusters corresponding to coarse-grained topics. We evaluate the computational efficacy of the proposed method and demonstrate its systematic improvement in scalability as the data volume increases. Next, we apply the proposed method to large-scale Twitter data (26 million tweets) about the COVID-19 Vaccination in Japan. The proposed method separately identifies the reactions to news and the reactions to tweets.

We study node similarity in labeled networks, using the label sequences found in paths of bounded length q leading to the nodes. (This recalls the q-grams employed in document resemblance, based on the Jaccard distance.) When applied to networks, the challenge is two-fold: the number of q-grams generated from labeled paths grows exponentially with q, and their frequency should be taken into account: this leads to a variation of the Jaccard index known as Bray-Curtis index for multisets. We describe nSimGram, a suite of fast algorithms for node similarity with q-grams, based on a novel blend of color coding, probabilistic counting, sketches, and string algorithms, where the universe of elements to sample is exponential. We provide experimental evidence that our measure is effective and our running times scale to deal with large real-world networks.

We address the problem of un-supervised soft-clustering that we call micro-clustering. The aim of the problem is to enumerate all groups composed of records strongly related to each other, whereas standard clustering methods find boundaries at which records are few. The existing methods have several weak points; generation of intractable amounts of clusters, biased size distributions, lack of robustness, etc. We propose a new methodology data polishing. Data polishing clarifies the cluster structures in the data by perturbating the data according to feasible hypothesis. More precisely, for graph clustering problems, data polishing replaces dense subgraphs that would correspond to clusters by cliques, and deletes edges not included in any dense subgraph. The clusters are clarified as maximal cliques, thus are easy to find, and the number of maximal cliques is reduced to tractable numbers. We also propose an efficient algorithm so that the computation is done in few minutes even for large scale data. The computational experiments demonstrate the efficiency of our formulation and algorithm, i.e., the number of solutions is small, such as 1,000, the members of each group are deeply related, and the computation time is short.

In this paper, we propose a first application of data mining techniques to propositional satisfiability. Our proposed mining based compression approach aims to discover and to exploit hidden structural knowledge for reducing the size of propositional formulae in conjunctive normal form (CNF). It combines both frequent itemset mining techniques and Tseitin's encoding for a compact representation of CNF formulae. The experimental evaluation of our approach shows interesting reductions of the sizes of many application instances taken from the last SAT competitions.

We propose a new data mining method called frequentness-transitional pattern mining for finding patterns with interesting sequential behavior specified by a user's query. For a series of databases, we introduce the frequentness-sequence of a pattern that is a sequence of the two symbols ‘H’ and ‘L,’ which represent the frequency or infrequency in each segment of a database, respectively. The problem is finding patterns whose frequentness-sequences satisfy the query. The goal of this research is to develop an efficient algorithm and its implementation that accepts various models and that can be widely used in practice with large-scale data. Thus, we chose an itemset as a pattern, and regular expression for the query language to accept various models. To cope with the unavoidably large number of candidate patterns, we use Zero-suppressed Binary Decision Diagrams (ZDDs or ZBDDs) to store and operate a large number of candidate itemsets in a short time. Our algorithm performed quite well in our computational experiments, such that it is competitive with the standard itemset mining algorithms that can be used only to find frequent itemsets. To the best of our knowledge, this is the first study on detecting distinctive itemsets of user-specific models of sequential behaviors.

In this paper, we study efficient closed pattern mining in a general framework of set systems, which are families of subsets ordered by set-inclusion with a certain structure, proposed by Boley, Horváth, Poigné, Wrobel (PKDD'07 and MLG'07). By modeling semi-structured data such as sequences, graphs, and pictures in a set system, we systematically study efficient mining of closed patterns. For a class of accessible set systems with a tree-like structure, we present an efficient depth-first search algorithm that finds all closed sets in accessible set systems without duplicates in polynomial-delay and polynomial-space w.r.t. the total input size using efficient oracles for the membership test and the closure computation for the pattern class. From the above results, we show that the closed pattern mining problems are efficiently solvable both in time and space for the following classes: convex hulls, picture patterns in 2-D planes, maximal bi-cliques, closed relational graphs, closed patterns for rigid motifs with wildcards.