Making sense of gigabytes: a system for knowledge-based market analysis
Tej Anand, Gary S. Kahn · Innovative Applications of Artificial Intelligence · 1992
Market researchers are looking more and more to point-of-sale scanners in retail outlets as a rich source of high-quality and timely sales data. Unfortunately, the sheer volume of such data makes analysis and interpretation an overwhelming task. Consequently, there is a demand for software tools that can automatically provide an analysis of large volumes of data. SPOTLIGHT is a knowledge-based product that enables A. C. Nielsen clients to understand what is significant in databases of point-of-sale scanner data. Using SPOTLIGHT, manufacturers and retailers track the sale and movement of products, assess the effectiveness of various promotional strategies, and compare the performance of competing products and product segments. SPOTLIGHT provides an alternative to the classical approach, where market analysts provide custom interpretations and reports, typically using spreadsheet tools such as Lotus as aids. SPOTLIGHT turns a task that takes 2 to 4 weeks into a task of 15 minutes to several hours depending on data volumes. Although previous attempts were made to automate the analysis of market data, within this domain, SPOTLIGHT represents the first commercial use of an expert system shell and the first deployment of sophisticated analytic capabilities directly onto a large number of widely distributed personal computer (PC) platforms.