Crowdsourcing represents a powerful approach that seeks to harness the collective knowledge or creativity of a large and independent network of people for organizations. While the approach drastically facilitates the sourcing and aggregating of information, it represents a latent challenge for organizations to process and evaluate the vast amount of crowdsourced contributions – especially when they are submitted in an unstructured, textual format. In this study, we present an on-going design science research project that is concerned with the construction of a design theory for semi-automated information processing and decision support in crowdsourcing. The proposed concept leverages the power of crowdsourcing in combination with text mining and machine learning algorithms to make the evaluation of textual contributions more efficient and effective for decision-makers. Our work aims to pro-vide the theoretical foundation for designing such systems in crowdsourcing. It is intended to contribute to decision support and business analytics research by outlining the capabilities of text mining and machine learning techniques in contexts that face large amounts of user-generated content. For practitioners, we provide a set of generalized design principles and design features for the implementation of these algorithms on crowdsourcing platforms.

First Page


Last Page