Big data in fashion industry: Color cycle mining from runway data

Color is a powerful selling tool, especially in the fashion and textile industry, in which products aim to inspire consumers visually. Color Cycle Analysis studies the recurring cycle of trends. Traditional fashion color cycle analysis and prediction is performed by observing and extrapolating from trends apparent on fashion runways. With the emergence of big data, there is a potential to apply data analytics method in fashion industry. We propose and develop a data-driven methodology to analyze color trends by mining online textual data of global fashion runways collected from the Style.com website. By capturing three important elements in color hue, saturation and brightness, we are able effectively extract their presence and variations in textual data. We illustrate the reoccurrence of seven Color Cycle phases: High Chroma, Multicolored, Subdued, Earth Tones, Achromatic, and Purple Phase from runway review data.