Definition of customer requirements, function requirements, and product structures based on big data and data mining methods
| dc.contributor.author | Shi, Yanlin | |
| dc.contributor.examiningcommittee | Liang, Xihui (Mechanical Engineering) | en_US |
| dc.contributor.examiningcommittee | Thulasiram, Ruppa (Computer Science) | en_US |
| dc.contributor.examiningcommittee | Xue, Deyi (Mechanical and Manufacturing Engineering, University of Calgary) | en_US |
| dc.contributor.supervisor | Peng, Qingjin | |
| dc.date.accessioned | 2022-03-18T16:46:45Z | |
| dc.date.available | 2022-03-18T16:46:45Z | |
| dc.date.copyright | 2022-03-18 | |
| dc.date.issued | 2022-03-14 | |
| dc.date.submitted | 2022-03-14T20:15:19Z | en_US |
| dc.date.submitted | 2022-03-18T16:36:33Z | en_US |
| dc.degree.discipline | Mechanical Engineering | en_US |
| dc.degree.level | Doctor of Philosophy (Ph.D.) | en_US |
| dc.description.abstract | Definition of customer requirements (CRs), function requirements (FRs), product structures (PSs) has significant impacts on product development. Traditional methods of defining CRs, FRs and PSs such as quality function deployment (QFD), expert evaluation and benchmarking highly rely on experience of experts and designers. This research proposes methods based on big data and data mining to improve accuracy and efficiency of decision-making in defining CRs, FRs, and PSs. Online product customer comments are searched by a focused crawling method. Collected customer comments are filtered based on parts of speech and frequency of words. Filtered data are clustered into groups by an affinity propagation (AP) clustering method to define CRs. Importance rates (IRs) of CRs are then decided by integration of the importance-performance analysis and Kano model to balance conflict comments from different customers using a similarity matrix in the spectral clustering method. FRs of a product are defined based on the function description of online products crawled using the focused crawling method. Minimum and maximum FR implementations of the product are decided by polynomial modeling and least square methods. IRs of FRs are defined by adjusting initial weight of FRs using the information entropy based on defined IRs of CRs. PSs are then defined based on relations of FRs and physical components in benchmarking products collected from online websites using WordNet hierarchy and association relation methods. By comparing performance of PSs using IRs of FRs and relations of FRs and PSs, the best PSs for meeting each FR are selected from potential design solutions using a pairwise comparison method. The proposed methods are applied in design of upper limb rehabilitation devices to improve accuracy and efficiency in definitions of CRs, IRs of CRs, FRs implementation, IRs of FRs, and PSs. Results show that the methods can significantly improve quality of concept design in definitions of CRs, FRs, and PSs of product. | en_US |
| dc.description.note | May 2022 | en_US |
| dc.identifier.uri | http://hdl.handle.net/1993/36367 | |
| dc.language.iso | eng | en_US |
| dc.rights | open access | en_US |
| dc.subject | customer requirements | en_US |
| dc.subject | function requirements | en_US |
| dc.subject | product structures | en_US |
| dc.subject | big data | en_US |
| dc.subject | data mining | en_US |
| dc.title | Definition of customer requirements, function requirements, and product structures based on big data and data mining methods | en_US |
| dc.type | doctoral thesis | en_US |
| oaire.citation.endPage | 1291 | en_US |
| oaire.citation.issue | 5 | en_US |
| oaire.citation.startPage | 1279 | en_US |
| oaire.citation.title | Journal of Process Mechanical Engineering | en_US |
| oaire.citation.volume | 235 | en_US |
| project.funder.name | Natural Sciences and Engineering Research Council (NSERC) of Canada | en_US |