Please use this identifier to cite or link to this item: http://hdl.handle.net/10603/234503
Title: Efficient and Effective Constraint based Sequential Pattern Mining SPM algorithm for understanding the customers buying behaviour from Time stamp based Sequence Dataset
Researcher: Desai Niti Ashishkumar
Guide(s): Ganatra Amit P.
Keywords: Constraint based Sequential Pattern Mining
Data Mining
Emerging Patterns
University: Uka Tarsadia University
Completed Date: 2017
Abstract: The main objective of an economic or business activity is to satisfy needs and wants of customers in order to generate profits. Business Managers seek to identify and predict purchase tendency of customers, with a view to plan their business strategies, including product development and marketing sub-strategies. Various data mining techniques including Sequential Pattern Mining (SPM) techniques help in achieving aforesaid objective. newlineOur research is focused on purchase trend of customer, where timing of purchase is more important than association of items to be purchased, and which can be found out with Sequential Pattern Mining methods. Conventional Apriori and FP-Growth based sequential pattern mining algorithms have worked purely on frequency parameter (support and confidence framework). Existing algorithms identify patterns that were more frequent but suffer from challenges like generation of huge number of patterns, lack of user?s interested patterns, rare item problem etc. Incorporation of six additional constraints like Gap/Duration, Compactness, Item, Recency, Profitability and Length along with conventional Frequency is able to address such shortcomings. newlineConstraint Sequential Pattern Mining framework is to ensure that all patterns are recently active, active for certain time span, profitable and indicative of next timeline for purchase. From the experiment study on 13 various standard secondary and real time datasets, it has been revealed that incorporation of seven constraints in one algorithm improves execution time by average 4.59% w.r.t Prefixspan and 3.11% w.r.t. RFM algorithm. Further Emerging Patterns that can help in predicting future buying behaviour can also be identified with the Constraint based Emerging Pattern framework. Existing Recommender System predicts few item(s) that customers are more likely to purchase in future. Constraint based recommender system, furthermore focuses on recommendations to be made for items of purchase in near and long term future, based on incorporation of Gap/...
Pagination: All Pages
URI: http://hdl.handle.net/10603/234503
Appears in Departments:Faculty of Engineering and Technology

Files in This Item:
File Description SizeFormat 
01_title.pdfAttached File102.06 kBAdobe PDFView/Open
02_certificate.pdf808.43 kBAdobe PDFView/Open
03_preliminary.pdf606.89 kBAdobe PDFView/Open
04_chapter 1.pdf254.34 kBAdobe PDFView/Open
05_chapter 2.pdf598.13 kBAdobe PDFView/Open
06_chapter 3.pdf557.52 kBAdobe PDFView/Open
07_chapter 4.pdf516.32 kBAdobe PDFView/Open
08_chapter 5.pdf1.24 MBAdobe PDFView/Open
09_chapter 6.pdf2.02 MBAdobe PDFView/Open
10_chapter 7.pdf347.47 kBAdobe PDFView/Open
11_chapter 8.pdf123.7 kBAdobe PDFView/Open
11_publications.pdf120.39 kBAdobe PDFView/Open
11_references.pdf468.34 kBAdobe PDFView/Open
12_glossary.pdf130.59 kBAdobe PDFView/Open


Items in Shodhganga are protected by copyright, with all rights reserved, unless otherwise indicated.

Altmetric Badge: