The "business environment virtualization" software NEC is currently developing may masquerade as a way for companies to keep track of their competitors online (see pie charts, etc), but we here can see it for what it really is: a scheme to create an unlimited army of lifeless tech-blogging drones. With its "cruising of RSS and web pages" for "product features" and algorithmically processing press releases for "conducting a comparative analysis of products," this software could probably crank out some pretty solid posts when it wasn't working to crush its master's rival companies, all without the requisite Mountain Dew and Cheetos breaks. We're not too worried here, though. We know you guys come for the gadgets, but stay for the red-blooded passion, right? Some relevant bits from the press release after the jump. Oh the irony!
(1) Industry Structural Analysis Industry-based information tends to be randomly delivered by many companies in a variety of different formations, including RSS and web pages. Analysis of this information requires the combination of these formats. "Industry structural analysis technology" assesses the relationship between seller and buyer for raw materials and finished products using algorithms that integrate information from different formats and evaluates this information in light of analysis results from previous corporate activities. Based on this assessment, regular analysis is conducted to determine the company's position and its relationship with other companies in terms of products and services.
(2) Comparative Product Analysis
The presentation of product information released on web pages differs from company to company. It is therefore necessary to understand the meaning of each presentation and classify it accurately when conducting a comparative analysis of products. "Comparative analysis technology" for comparing rival products determines the characteristic differences between products by looking at product information spread across multiple web pages, as well as product outlines presented in different formats. This is done by applying an algorithm to identify product pages by scanning for keywords such as "outline" and "features" as well as relationships between web pages. Another algorithm is then used to calculate the similarities between product features by scanning for keywords such as "high resolution" and "thinnest".
[via Digital World Tokyo]