Please use this identifier to cite or link to this item: https://hdl.handle.net/10356/150900
Title: Knowledge graph-based explainable knowledge recommendation in product development
Authors: Huang, Weifeng
Keywords: Engineering::Mechanical engineering
Issue Date: 2021
Publisher: Nanyang Technological University
Source: Huang, W. (2021). Knowledge graph-based explainable knowledge recommendation in product development. Final Year Project (FYP), Nanyang Technological University, Singapore. https://hdl.handle.net/10356/150900
Abstract: Product development nowadays has become more challenging than ever. Enormous heterogeneous knowledge is always involved in the process forcing the decision-making much complicated. However, most of the exiting decision-making tool like conventional knowledge recommendation system is not able to provide explicit explanation for its recommendation result, which significantly jeopardizes the persuasibility and reliability of the system. Hence, this project aims to fill the gap by establishing an explainable knowledge graph-based knowledge recommendation system by path extraction. A case study using 3D printing troubleshooting dataset is conducted to demonstrate the application of the established knowledge recommendation system in product development scenario. Moreover, two knowledge recommendation approaches: TF-IDF-based cosine similarity as the baseline and a more advanced embedding-based recommendation system (RS) called Ripplenet were conducted via the same dataset and compared with the established system in discussion section.
URI: https://hdl.handle.net/10356/150900
Fulltext Permission: restricted
Fulltext Availability: With Fulltext
Appears in Collections:MAE Student Reports (FYP/IA/PA/PI)

Files in This Item:
File Description SizeFormat 
Knowledge graph-based explainable knowledge recommendation in product development.pdf
  Restricted Access
2.07 MBAdobe PDFView/Open

Page view(s)

57
Updated on Dec 5, 2021

Download(s)

2
Updated on Dec 5, 2021

Google ScholarTM

Check

Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.