Please use this identifier to cite or link to this item:
http://hdl.handle.net/10603/325092
Title: | Concept Hierarchy Based Diverse Frequent Patterns |
Researcher: | M Kumara Swamy |
Guide(s): | P Krishna Reddy |
Keywords: | Computer Science Computer Science Information Systems Engineering and Technology |
University: | International Institute of Information Technology, Hyderabad |
Completed Date: | 2019 |
Abstract: | As a part of data mining, several approaches have been developed for extracting different kinds of interesting knowledge from the data. In this thesis, we propose a model of diverse frequent patterns (DFPs) and show that it can be employed to improve the performance of RS and QR. newlineWe propose a model to compute the diversity of pattern (set of items). Given a domain, set of items can be grouped into a category, and a pattern may contain the items which belong to multiple categories. We propose a framework to rank the pattern by analyzing the extent the items of the pattern belong to different categories in the corresponding concept hierarchy (CH). We defined a diversity measure to rank the patterns called drank. A pattern which satisfies a threshold of drank is called a diverse pattern (DP). A pattern which satisfies a threshold of drank and support are called diverse frequent pattern (DFP). We propose two frameworks to compute the diversity of the pattern: based on balanced CH and unbalanced CH. Through experimental results, we show that the DFPs generate different knowledge to FPs. newlineExploiting DFP, we proposed an improved association rule-based RS (ARS) by observing high accuracy of recommendation alone cannot meet user satisfaction. The ARS is aimed to improve the accuracy and diversity. The experimental results on MovieLens dataset show that the proposed ARS improves the performance of the existing ARS with better diversity without compromising the accuracy. newlineFurther, by exploiting DFP, we have proposed an approach for search QRs. Search engines use a QR to recommend a different set of queries to improve the user satisfaction. We propose an approach for QR by extending the concepts of ARs and DFPs, and unbalanced CH of search terms. The experimental results on AOL dataset show that QRs with the proposed approach improves the diversity without compromising accuracy. newlineOverall, we have proposed a new model of DFPs, and demonstrated the utility DFPs in improving the performance of AR based RS and search QRs |
Pagination: | |
URI: | http://hdl.handle.net/10603/325092 |
Appears in Departments: | Computer Science and Engineering |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
80_recommendation.pdf | Attached File | 1.37 MB | Adobe PDF | View/Open |
appendix.pdf | 806 kB | Adobe PDF | View/Open | |
certificate.pdf | 40.9 kB | Adobe PDF | View/Open | |
chapter 1.pdf | 2.21 MB | Adobe PDF | View/Open | |
chapter 2.pdf | 4.58 MB | Adobe PDF | View/Open | |
chapter 3.pdf | 6.82 MB | Adobe PDF | View/Open | |
chapter 4.pdf | 5.1 MB | Adobe PDF | View/Open | |
chapter 5.pdf | 6.35 MB | Adobe PDF | View/Open | |
title.pdf | 74.54 kB | Adobe PDF | View/Open |
Items in Shodhganga are licensed under Creative Commons Licence Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0).
Altmetric Badge: