Mining approximate multi-relational patterns

Abstract

Abstract-Three recent trends aim to make local pattern mining more directly suited for use on data as it presents itself in practice, namely in a multi-relational form and affected by noise. The first of these trends is the generalisation of local pattern syntaxes to approximate, noise-tolerant, variants (notably faulttolerant itemset mining and community detection). The second of these trends is to develop pattern syntaxes that are directly applicable to multi-relational data. The third one is to better quantify the interestingness of and redundancy between such local patterns.In this paper we leverage recent results from these lines of research to introduce a noise-tolerant pattern syntax for multirelational data. We show how enumerating all patterns of this syntax in a given database can be done remarkably efficiently. We contribute a way to quantify the interestingness of these patterns, thus overcoming the pattern explosion problem. And finally, we show the usefulness of the pattern syntax and the scalability of the algorithm by presenting experimental results on real world and synthetic data.

P-N-RMiner: a generic framework for mining interesting structured relational patterns

“…Ideally, the pattern syntax is tolerant to missing relations to ensure noise resilience, similar to [22]. The interestingness can be made more versatile by considering a more varied range of prior belief types.…”

“…Like previous work on mining interesting patterns in relational data [22][23][24], our algorithm is based on the fixpoint-enumeration algorithm by Boley et al [1]. Although that algorithm already exists, it should be noted that it is a meta-algorithm, which does not directly work on the data.…”

“…This algorithmic scheme has been used before in the data mining literature for enumerating maximal CCSs [24], including extensions to n-ary relations [23] and approximate CCSs [22]. Here, we adhere to the same algorithmic scheme.…”

“…One could also consider approximate patterns by discarding the completeness constraint. This would lead to an increased computational complexity, but the increase has been shown to be manageable [22]. For simplicity, we do not consider approximate patterns in this paper.…”

“…From this we extracted the relationship between books and categories and the hierarchy itself, keeping two levels below the category Book → Subject. The dataset that we obtain has the structure shown in Figure 4 and consists of 22,003 Rating Book Customer Subject Fig. 4 Relational schema of the Amazon Book Ratings data books, 9,855 customers, 417 hierarchically structured book subjects, as well as 36,415 ratings and 53,403 subject memberships.…”