Multidimensional Association Rules in Boolean Tensors

  title={Multidimensional Association Rules in Boolean Tensors},
  author={Kim-Ngan Nguyen and Lo{\"i}c Cerf and Marc Plantevit and Jean-François Boulicaut},
Popular data mining methods support knowledge discovery from patterns that hold in binary relations. We study the generalization of association rule mining within arbitrary n-ary relations and thus Boolean tensors instead of Boolean matrices. Indeed, many datasets of interest correspond to relations whose number of dimensions is greater or equal to 3. However, just a few proposals deal with rule discovery when both the head and the body can involve subsets of any dimensions. A challenging… 

Figures and Tables from this paper

Mining Disjunctive Rules in Dynamic Graphs

This work builds upon a formalization of association rules that hold in n-ary Boolean tensors to design more expressive local patterns in this special case of dynamic graph where the set of vertices remains unchanged though edges that connect them may appear or disappear at the different timestamps.

Discovering descriptive rules in relational dynamic graphs

This work designs the pattern domain of multi-dimensional association rules, i.e., non trivial extensions of the popular association rules that may involve subsets of any dimensions in their antecedents and their consequents and proposes optimizations to support both rule extraction scalability and non redundancy.

Multidimensional Data Mining Based on Tensor

By using theTDM, high order data structure and naive description for information retrieval are possible by using the TDM, and several operations specific to multidimensional data mining to reduce amount of description can be modeled by using TDM.

Multidimensional Data Mining Based on Tensor Model

  • Ryohei YokobayashiT. Miura
  • Computer Science
    2018 IEEE First International Conference on Artificial Intelligence and Knowledge Engineering (AIKE)
  • 2018
This work shows TDM allows us to model several operations specific to multidimensional data mining yet to reduce amount of description, and discusses multiddimensional rule mining here.

Finding maximal homogeneous clique sets

This work proposes to look for patterns called maximal homogeneous clique sets, where such a pattern is a subgraph that is structured in several large cliques and where all vertices share enough labels, and presents an algorithm based on graph enumeration to compute all patterns satisfying user-defined constraints.

New Applications of Formal Concept Analysis : A Need for Original Pattern Domains

This work surveys the results obtained by the research group concerning the design of pattern domains to support knowledge discovery and information retrieval in arbitrary n-ary relations and focuses on a real data mining perspective.

New applications of Formal Concept Analysis: a Need for New Pattern Domains

This work surveys the results obtained by the research group concerning the design of pattern domains to support knowledge discovery and information retrieval in arbitrary n-ary relations and focuses on a real data mining perspective.

Co-evolution pattern mining in dynamic attributed graphs. (Fouille de motifs de co-evolution dans des graphes dynamiques attribués)

This thesis proposes a new pattern domain that has been called co-evolution patterns, and proposes three constraint-based algorithms, called MINTAG, H-MINTAG and Sky-H-MintAG, that are complete to extract the set of all patterns that meet the different constraints.

Finding lasting dense subgraphs

This paper provides definitions for density over multiple graph snapshots, that capture different semantics of connectedness over time, and proposes a set of efficient algorithms to solve the Best Friends Forever problem.

Cohesive Co-evolution Patterns in Dynamic Attributed Graphs

This work designs the first algorithm to mine the complete set of cohesive co-evolution patterns in a dynamic graph and demonstrates that the algorithm enables to discover relevant patterns inA feasible time.



Closed patterns meet n-ary relations

Considering the special case of ternary relations, Data-Peeler outperforms the state-of-the-art algorithms CubeMiner and Trias by orders of magnitude and a collection of closed n-sets is shown to be an excellent starting point to compute a tiling of the dataset.

Cubegrades: Generalizing Association Rules

An important task for computing cubegrades is described: generation of significant cubes which is analogous to generating frequent sets, and a novel Grid Based Pruning (GBP) method is employed for this purpose.

Multiple Uses of Frequent Sets and Condensed Representations (Extended Abstract)

This paper shows how frequent sets can be used as a condensed representation for answering various types of queries, and defines a general notion of condensed representations, and shows that frequent sets, samples and the data cube can be viewed as instantations of this concept.

Mining Positive and Negative Association Rules: An Approach for Confined Rules

An algorithm is proposed that extends the support-confidence framework with sliding correlation coefficient threshold and discovers negative association rules with strong negative correlation between the antecedents and consequents.

Mining all frequent projection-selection queries from a relational table

The main contribution of this paper is the proposal of a level-wise algorithm for mining all frequent projection-selection queries in a data warehouse over a star schema, and it is shown that the complexity in the number of scans of the algorithm is similar to that of the well known Apriori algorithm.

Discovery of frequent DATALOG patterns

WARMR is presented, a general purpose inductive logic programming algorithm that addresses frequent query discovery: a very general DATALOG formulation of the frequent pattern discovery problem.

Mining Graph Evolution Rules

GERM (Graph Evolution Rule Miner), an algorithm to mine all graph-evolution rules whose support and confidence are greater than given thresholds, is applied to analyze four large real-world networks, suggesting the usage of these local patterns to globally discriminate different kind of networks.

Mining Association Rules in Multiple Relations

The system Warmr is presented, which extends Apriori to mine association rules in multiple relations, and is applied to the natural language processing task of mining part-of-speech tagging rules in a large corpus of English.

A template model for multidimensional inter-transactional association rules

This study presents a template model to help users declare the interesting multidimensional inter-transactional associations to be mined and shows, through a series of experiments on both synthetic and real-life data sets, that these optimization techniques can yield significant performance benefits.