CMRules: Mining sequential rules common to several sequences

Abstract

Abstract-Sequential rule mining is an important data mining task with wide applications. However, current algorithms for discovering sequential rules common to several sequences use very restrictive definitions of sequential rules, which make them unable to recognize that similar rules can describe a same phenomenon. This can have many undesirable effects such as (1) similar rules that are rated differently, (2) rules that are not found because they are considered uninteresting when taken individually, (3) and rules that are too specific, which makes them less likely to be used for making predictions. In this paper, we address these problems by proposing a more general form of sequential rules such that items in the antecedent and in the consequent of each rule are unordered. We propose an algorithm named CMRules, for mining them. The algorithm proceeds by first finding association rules to prune the search space for items that occur jointly in many sequences. Then it eliminates association rules that do not meet the minimum confidence and support thresholds according to the time ordering. We evaluate the performance of CMRules in three different ways.First, we provide an analysis of its time complexity. Second, we compare its performance (execution time, memory usage and scalability) with an adaptation of an algorithm from the literature that we name CMDeo. For this comparison, three real-life public datasets representing three types of data and having varied characteristics. Results show that for some datasets CMRules is faster and has a better scalability for low support thresholds. Lastly, we report a successful application of the algorithm in a tutoring agent.

Mining association rules for the quality improvement of the production process

“…A subset of the problem of association rule mining is the problem of mining sequential rules common to several sequences as follows (Fournier-Viger, Faghihi, Nkambou, & Mephu Nguifo, 2012). A sequence database SD is a set of sequences S = {s 1 , s 2 , .…”

“…Especially, for the problem of mining sequential rules common to several sequences, the Pattern-Growth approach could be particularly valuable in managing complex tasks such as monitoring the state and quality of materials resources in industrial operational processes. For that reason, we use the RuleGrowth algorithm (Fournier-Viger, Nkambou, & Tseng, 2011) relying on a PatternGrowth, in order to discover a more general form of sequential rules such that items in the antecedent and in the consequent of each rule are unordered. This form of sequential rules conveys more information and it is not discovered by other approaches stating that items of the left part or the right part of a rule have to appear with exactly the same ordering in a sequence (Lo, Khoo, & Wong, 2009).…”

A Review of Modern Sequential Rule Mining Techniques

“…CMRules: an association rule mining based mostly algorithm for the invention of sequential rules [24].…”

Improving SOA antipatterns detection in Service Based Systems by mining execution traces

“…Since sequential patterns are more general than the pure sequence ones, mining algorithms designed for the former might prove to be less efficient when applied to the latter (as additional steps might be required for listing all significant sets). Nevertheless, to jump-start our experimental study and given the specificity of our datasets, we choose the RuleGrowth algorithm [31] that seemed to fit at best. Although it has not been optimized for pure sequences its performances are more than satisfactory.…”