Journal of Fashion Marketing and Management · 2007 · 69 citations · 16 references
BusinessCommunicationMarketingDemand ForecastingDigital MarketingForecastingFashionFashion MarketingTrend DetectionData-driven Fashion ForecastingSocial Media Trend AnalysisStreet FashionTrend Classification
Current fashion forecasting methods lag behind market dynamics because they cannot keep pace with the rapid diffusion of data and information. This study investigates how weblog text mining can be used for fashion forecasting, particularly street fashion trends, to add a new forecasting dimension for the multi‑channel fashion market. The authors compare real‑time blog data with historical records and illustrate the approach using Nielsen BuzzMetrics' BlogPulse to detect lingering, changing, or dynamic trends. Successful fashion weblog forecasting requires a rich, structured archive of fashion blogs, systematic classification of industry web text, and long‑term research traditions to capture cultural, behavioral, linguistic, socioeconomic, and communication patterns, demonstrating the method’s applicability to the evolving fashion business.
Abstract Purpose – The purpose of this paper is to examine the theoretical/conceptual development and application of weblog‐textmining to fashion forecasting in general and street fashion trending in particular. Design/methodology/approach – The current methods of forecasting cannot keep pace with the changing dynamics of the marketplace – mostly due to the rampant diffusion of data/information. The company that can tap the continual flow of data/information in the present, contrast it with a stored set of information from the past, and adjust based on repeated cycles, will have the best insight into the lingering trend, changing trend, or dynamic trend. The paper uses a simple example to explain blog trend analysis using Nielsen BuzzMetrics' BlogPulse. Findings – The study finds that to make fashion weblog forecasting a reality, there needs to be a rich accumulation of fashion communication in structured blogs. In addition, there needs to be a classification of the various forms of industry web text, web venue. Furthermore, rich research traditions must be in place to chronicle the cultural, behavioral, linguistic, socioeconomic, and communication behaviors over time for the weblog and the fashion weblogger in particular. Practical implications – The changing dynamics of the fashion business makes it a good example for understanding the weblog‐text mining approach developed in this paper. Originality/value – The understanding and implementation of trend forecasting using blogs as data mining sources will add another dimension of forecasting techniques to survive the multi‐channel revolution in fashion marketing.
16
Information diffusion through blogspace
Daniel Gruhl, David Liben‐Nowell, R. Guha et al. · ACM SIGKDD Explorations Newsletter · 2004 · 951 citations
On the bursty evolution of blogspace
Ravi Kumar, Jasmine Novak, Prabhakar Raghavan et al. · 2003 · 562 citations
Do you want to keep your customers forever?
Long Range Planning · 1995 · 472 citations