NOSEP: Nonoverlapping Sequence Pattern Mining With Gap Constraints

Abstract

Sequence pattern mining aims to discover frequent subsequences as patterns in a single sequence or a sequence database. By combining gap constraints (or flexible wildcards), users can specify special characteristics of the patterns and discover meaningful subsequences suitable for their own application domains, such as finding gene transcription sites from DNA sequences or discovering patterns for time series data classification. Due to the inherent complexity of sequence patterns, including the exponential candidate space with respect to pattern letters and gap constraints, to date, existing sequence pattern mining methods are either incomplete or do not support the Apriori property because the support ratio of a pattern may be greater than that of its subpatterns. Most importantly, patterns discovered by these methods are either too restrictive or too general and cannot represent underlying meaningful knowledge in the sequences. In this paper, we focus on a nonoverlapping sequence pattern mining task with gap constraints, where a nonoverlapping sequence pattern allows sequence letters to be flexibly and maximally utilized for pattern discovery. A new Apriori-based nonoverlapping sequence pattern mining algorithm, NOSEP, is proposed. NOSEP is a complete pattern mining algorithm, which uses a specially designed data structure, Nettree, to calculate the exact occurrence of a pattern in the sequence. Experimental results and comparisons on biology DNA sequences, time series data, and Gazelle datasets demonstrate the efficiency of the proposed algorithm and the uniqueness of nonoverlapping sequence patterns compared to other methods.

NetNCSP: Nonoverlapping closed sequential pattern mining

“…As can be seen from Table 1, Reference [34] is the closest to this paper. The differences between Reference [34] and this paper are as follows.…”

“…This method aims to discover subsequences from sequences that satisfy the gap constraints and support threshold. Existing studies are based on three types of conditions, no-condition [29,56], oneoff condition [8,57], and nonoverlapping condition [33,34,58]. The no-condition allows sequence letters to be reused by patterns for an unlimited time.…”

“…The nonoverlapping condition allows the sequence letters to be rematched by the different pattern letters and to be matched no more than once by the same pattern letter. Although the nonoverlapping condition is more complex than the other two, our previous studies have demonstrated that the pattern mining under the nonoverlapping condition is not only a complete mining method with the Apriori property but can also discover more valuable patterns than the other two conditions [34,35]. Consequently, pattern mining under the nonoverlapping condition outperforms those of the no-condition and one-off condition.…”

“…Introducing the nonoverlapping condition not only reduces the number of occurrences but also makes the unique patterns richer. Our previous studies have proved that the nonoverlapping SPM is a complete mining method which satisfies the Apriori property [34,35]. Example 1 is used to explain the periodic gap pattern, gap constraint mining, and nonoverlapping support.…”

“…Our previous work proposed an effective algorithm NOSEP and reported that the nonoverlapping SPM has better performance than other state-of-the-art gap constraint SPM methods in finding useful patterns in biology sequences and avoiding under-expression and over-expression in time series [34]. However, all frequent patterns discovered by NOSEP can be furtherly compressed.…”