A Knowledge-Based System for Fashion Trend Forecasting

In this paper, we show how artificial intelligence techniques can be applied for the forecasting of trends in the high creative domain of fashion. We describe a knowledge-based system that, starting from a set of keywords and pictures representing the concepts on which a fashion stylist chooses to base a new collection, is able to automatically create a trend forecast composed by the set of colors that better express these target concepts. In order to model the knowledge used by the system to forecast trends, we experimented Bayesian networks. This kind of model is learned from a dataset of past trends by using different algorithms. We show how Bayesian networks can be used to make the forecast and the experiments made in order to evaluate their performances.

[1]  Nada Lavrac,et al.  Machine Learning for Data Mining in Medicine , 1999, AIMDM.

[2]  Gregory F. Cooper,et al.  A Bayesian method for the induction of probabilistic networks from data , 1992, Machine Learning.

[3]  Ian H. Witten,et al.  Data mining: practical machine learning tools and techniques, 3rd Edition , 1999 .

[4]  Judea Pearl,et al.  Bayesian Networks , 1998, Encyclopedia of Social Network Analysis and Mining. 2nd Ed..

[5]  Margaret A. Boden.  Agents and creativity , 1994, CACM.

[6]  Harold Cohen.  The further exploits of Aaron, painter , 1995 .

[7]  M. Minsky.  The Society of Mind , 1986 .

[8]  Michael A. Arbib,et al.  The handbook of brain theory and neural networks , 1995, A Bradford book.

[9]  Douglas B. Lenat,et al.  Why AM and EURISKO Appear to Work , 1984, Artif. Intell..

[10]  J. Ross Quinlan,et al.  Induction of Decision Trees , 1986, Machine Learning.

[11]  Ian Witten,et al.  Data Mining , 2000 .