Product Classification - A Hierarchical Approach
Mikael K. Karlsson, Anton Karlstedt · Lund University Publications Student Papers (Lund University) · 2016
The social and environmental impact associated with consuming a product is something that is becoming increasingly important to consumers and businesses alike. This impact can in theory be computed simply by classifying a product and by mapping the classified product to the corresponding life-cycle assessments research. However, this type of mapping requires an extensive product taxonomy with a large plurality of categories, which makes classification through machine learning a non-trivial task. This thesis describes the implementation of a hierarchical product classifier using the Python library scikit-learn, that makes it possible to automatically classify products based primarily on their brands and titles into a large taxonomy. Using a data set of 3.1 million products spread over 800 categories, we trained a set of hierarchical classifiers using different learning algorithms. Evaluations of these algorithms showed that the best hierarchical classifier reached a hierarchical F1-score (hF1) of 0.85.