Research Papers

Trend Mining for Predictive Product Design

[+] Author and Article Information
Conrad S. Tucker

Mem. ASME 213-N Hammond Building, Engineering Design and Industrial Engineering,  The Pennsylvania State University, University Park, PA 16802ctucker4@psu.edu

Harrison M. Kim1

Mem. ASME 104 S. Mathews Avenue, Industrial and Enterprise Systems Engineering,  University of Illinois, Urbana-Champaign Urbana, IL 61801hmkim@illinois.edu


Corresponding author.

J. Mech. Des 133(11), 111008 (Nov 11, 2011) (11 pages) doi:10.1115/1.4004987 History: Received February 10, 2011; Revised August 10, 2011; Published November 11, 2011; Online November 11, 2011

The Preference Trend Mining (PTM) algorithm that is proposed in this work aims to address some fundamental challenges of current demand modeling techniques being employed in the product design community. The first contribution is a multistage predictive modeling approach that captures changes in consumer preferences (as they relate to product design) over time, hereby enabling design engineers to anticipate next generation product features before they become mainstream/unimportant. Because consumer preferences may exhibit monotonically increasing or decreasing, seasonal, or unobservable trends, we proposed employing a statistical trend detection technique to help detect time series attribute patterns. A time series exponential smoothing technique is then used to forecast future attribute trend patterns and generates a demand model that reflects emerging product preferences over time. The second contribution of this work is a novel classification scheme for attributes that have low predictive power and hence may be omitted from a predictive model. We propose classifying such attributes as either standard, nonstandard, or obsolete by assigning the appropriate classification based on the time series entropy values that an attribute exhibits. By modeling attribute irrelevance, design engineers can determine when to retire certain product features (deemed obsolete) or incorporate others into the actual product architecture (standard) while developing modules for those attributes exhibiting inconsistent patterns throughout time (nonstandard). Several time series data sets using publicly available data are used to validate the proposed preference trend mining model and compared it to traditional demand modeling techniques for predictive accuracy and ease of model generation.

Copyright © 2011 by American Society of Mechanical Engineers
Your Session has timed out. Please sign back in to continue.



Grahic Jump Location
Figure 1

Overall flow of preference trend mining methodology

Grahic Jump Location
Figure 2

Attribute-class distributions over time (attribute a1,1 is highlighted although both attribute patterns change over time)

Grahic Jump Location
Figure 3

Characterizing attribute preference trend over time

Grahic Jump Location
Figure 4

Example decision tree result for product design

Grahic Jump Location
Figure 5

Product design implications of attribute irrelevance classification

Grahic Jump Location
Figure 6

Attribute (Ai ) characterization (relevant and irrelevant categorization) from iteration 1 to iteration m (each iteration contains a total of n time series data sets)

Grahic Jump Location
Figure 7

Time series gain rRatio at iteration 1 (Period 1–12 with Period 13 predicted by employing the Holt-Winters predictive model)

Grahic Jump Location
Figure 8

Decision tree model using Period 12, 2009 data set only for model generation (results attained using Weka 3.6.1 [35])

Grahic Jump Location
Figure 9

Trend mining model using Periods 1–12, 2009 data for model generation (results attained using ESOL developed Java Based PTM code compatible with Weka) [35])

Grahic Jump Location
Figure 10

Time Series Attribute Entropy values for irrelevance characterization

Grahic Jump Location
Figure 11

Comparison of predictive accuracies between the PTM and DT models using 12 unseen time stamped data from 2010) [35])



Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In