Data Analytics for Creative Processes: Designing the Next Great Product
41 Pages Posted: 27 Jan 2020
Date Written: January 24, 2020
Abstract
Problem Definition: Over the past decades, firms that develop new flavor and fragrance products (a $30 billion per year market) have amassed large amounts of data from their own internal creation processes. Yet, new product creation still mostly relies on human expertise and arduous experimentation. We present a data analytics framework to aid in the flavor and fragrance creation process.
Practical Relevance: We describe general and fundamental algorithmic challenges for creating recipes. We address these challenges and propose an analytics framework to help product developers navigate the space of potential products that arises from combining ingredients and accelerate the overall development process.
Methodology: The framework consists of two steps. In the first step, we use data to fit a metric over products, a procedure we refer to as curation. The metric of choice is the Earth Mover's Distance (EMD), where distances between products are quantified in terms of the pairwise similarity between ingredients. The curation process is formulated as a metric learning problem, where the cost matrix used in EMD computations is directly fit from labeled data. In the second step, we develop analytical tools to aid in the design of new products by sampling and evaluating new products within an (EMD) neighborhood of a seed product, a procedure we refer to as creation.
Results: We introduce efficient algorithms for the curation and creation procedures. We derive parallel metric learning and k-NN search algorithms, an efficient sampling procedure, and an appropriate evaluation criterion across multiple metrics. We illustrate our results and product creation pipeline on open-access datasets.
Managerial Implications: The methods we propose are new formulations and algorithms that address challenges that we observed as common across the innovation process of different flavor and fragrance firms, and may serve as an inspiration for a broader computational creativity approach used in industry.
Keywords: Computational Creativity, Innovation Management, Metric Learning, Optimal Transport, Wasserstein Distance, Food Pairing
Suggested Citation: Suggested Citation