| dc.description.abstract | This thesis explored how artificial intelligence (AI), media platforms, and media users affect practices of fashion trend forecasting. Fashion trend forecasting is an essential part of the fashion system. Forecasters provide fashion companies with insights about emerging trends. New opportunities in the digital age transform forecasting practices: (1) AI offers opportunities to explore machine-made and data-driven fashion trend predictions. (2) Digital media platforms make available vast amounts of fashion-related content, which changes how forecasters collect information. (3) An increasing number of media users participate in fashion dissemination and thus contribute to the stream of information available to leverage. Such changes open up to new ways of shaping fashion trends. Current research offers limited knowledge about the broader context of digital development in the fashion industry to understand how the digital infrastructure, with AI, media platforms, and media users, reshape practices of fashion trend forecasting. The broad approach that includes three major aspects, AI, media platforms, and media users, was inspired by the analytical perspective of Andrew McAfee and Erik Brynjolfsson, which describe digital development focusing on machines, platforms, and the crowd. The topic was explored through a literature review about AI, media platforms, and media users, six in-depth interviews with professionals from the fashion and trend industries, and a qualitative content analysis of the websites of three trend forecasting agencies. The results of this thesis indicated that AI, though relatively unexplored, is increasingly relevant in fashion trend forecasting. AI is applicable in this sector mainly to observe emerging trends faster, to reduce overproduction, and to meet consumer needs more precisely. On the other hand, the results also indicated perceptions of AI as unfitted for some qualities of fashion, such as garment tactility and emotional values. Moreover, the results suggested both benefits and concerns regarding the use of content from digital fashion media in trend research. Some of the results indicated that popular social media provide insights about current trends but might not be the best to use in trend forecasting work as it is too present and mostly represents the mainstream. User participation in fashion media was highlighted in the results as a crucial factor that drives change in the fashion industry. Particularly concerning the role of consumers, which has become more visible and involved in the forming of trends. | eng |