A data mining approach to discovering reliable Incremental sequential patterns
| Author(s) | : | Nirali Parmar, Trupti Kodinariya |
| Institution | : | Computer Department, Gujarat Technological University Address |
| Published In | : | Vol. 2, Issue 14 — January 2015 |
| Page No. | : | - |
| Domain | : | Engineering |
| Type | : | Research Paper |
| ISSN (Online) | : | 2348-4470 |
| ISSN (Print) | : | 2348-6406 |
Sequential pattern mining is one of the data miningmethod for obtaining frequent sequential patterns in a sequentialdatabase. Generally sequential data mining methods could bedivided into two categories: Apriori-like methods and patterngrowth methods. In any sequential pattern, probability of timebetween two adjacent events could provide valuable informationfor decision-makers. As we know, there has been no methodologydeveloped to extract this probability in the sequential patternmining process. Here we extend the IncSpan algorithm andpropose a new sequential pattern mining approach: T-IncSpan.This approach imposes minimum time-probability constraint, sothat fewer but more reliable patterns will be obtained. T-IncSpanis compared with IncSpan in terms of number of patternsobtained and execution efficiency. Our experimental results showthat T-IncSpan is an efficient and scalable method for sequentialpattern mining.
Nirali Parmar, Trupti Kodinariya, “A data mining approach to discovering reliable Incremental sequential patterns”, International Journal of Advance Engineering and Research Development (IJAERD), Vol. 2, Issue 14, pp. -, January 2015.
