| dc.description.abstract | Artificial intelligence has become an essential component of modern industries, signif icantly influencing how businesses operate in the 21st century. The fashion industry,
known for its dynamic and fast-changing nature, is no exception. Fashion trends evolve
rapidly, with new styles spreading almost instantly through digital platforms. Social
media platforms such as Instagram, TikTok, Pinterest play a major role in shaping
consumer preferences, purchasing decisions, and trend adoption. Despite this shift,
many fashion brands still rely on traditional forecasting methods based on historical
sales data, seasonality, and expert judgment. These conventional approaches often
result in delayed responses to emerging trends, inaccurate demand forecasts, excess
inventory and limited design innovation. With the growing availability of real-time
data from social media interactions, online sales, there is a critical need for timely and
effective analytical solutions to process such data. Addressing this need, the present
study conducts a comprehensive review of AI applications used to enhance real-time
fashion trend forecasting by integrating social media and sales data. The review
examines various AI techniques employed in previous studies, explores data collection
and processing methods, evaluates the strengths and limitations of existing approaches.
It also identifies key challenges, including data inconsistency, lack of standardized
evaluation metrics, limited model transparency, and ethical concerns related to data
usage. Following internationally recognized review practices, a systematic literature
search was carried out using IEEE Xplore, ResearchGate, and Google Scholar, covering
studies published between 2013 and 2025. Overall, this review provides valuable
insights for fashion researchers, designers, retail strategists, highlighting research gaps
and supporting the development of more accurate, efficient, sustainable AI-driven
forecasting systems for the fashion industry. | en_US |
