The home textile market is subject to highly volatile fashion trends, exposing the manufacturers to substantial economic risks. They have to anticipate and adapt to constantly changing fashions, preferences and buyer behavior. Failures here lead to mistakes in product planning and production scheduling. As a consequence, currently up to 60 percent of the products turn out to be unsalable, while at the same time market opportunities are missed as attractive products were produced in too small volume. A major factor is a lack of communication and cooperation, due to a lack of effective networks among the manufacturers or with the large retailers, the only ones with immediate access to the customers.
In the AsIsKnown project we develop a semantic-based knowledge flow system for the European home textile industries. The system collects, across manufacturers and retailers, product information, monitors buying behavior in selected sales outlets and collects feedback on new products from interior designers and consumers. The manufacturers and retailers connected to the system can use the aggregated information to spot trends in the consumer behavior and to adapt and coordinate their product planning accordingly.
Technically, the AsIsKnown platform combines methods of text mining and data mining with Semantic Web technologies. The AsIsKnow web shop includes a virtual interior designer that generates suggestions tailored to the target group a user belongs to, and recommends products across manufacturers’ product lines. The program uses customer profiles that it aggregates from actual buying behavior in the web shop. Thus, the profiles are dynamically modified and kept up to date, reflecting the latest fashion trends.
The AsIsKnown system also 'reads' (via text mining) digitized textile and fashion magazines, looking for promising novel fashion ideas and documenting which materials and which colors are 'en vogue' in different contexts.