Brain Trust: Mining the Community Mind

Page 3 of 4


David Holtzman, CEO of Opion, Inc., a market forecasting company that analyzes online discussions, offers a different perspective on the use of community content: "Concordances are original works of art." Thus, Holtzman argues, the act of taking the unstructured data, such as message board content, and structuring it into a derivative proprietary format, involves some originality and is therefore in itself protected.

Opion aggregates public online discussions to determine trends and forecasts by looking at the community content and rating each author's influence using proprietary algorithms. Opion's practice is a blend of analysis and internal metrics. It begins by gathering the data from community spaces and then structures the data for analysis before providing the analyzed trend line information to subscribers of its services. The information can then be used internally against other data sources, such as sales and marketing information, to gauge the effectiveness of an advertising campaign or consumer sentiment on a brand.

According to its Web site, to date Opion has gathered data on over 557,000 identities and 11 million postings.

While there is very little argument that community content is publicly available, "anyone who looks at their server logs can tell if we are crawling their site," states Todd Bransford, director of product management for Cyveillance, Inc. "There have been instances where a site requests not to be crawled, and we always explain what it is we're doing, and in some cases put them on our exclude list."

"We serve the high-level executive who doesn't have the time to monitor or read the thousands of places where consumers may be talking about his or her company's products or services. Clients see us as an outsourced research center. Our value is in our technology that assists our skilled professional service individuals to analyze and verify the relevancy of the materials to the client before publishing research reports in a timely manner." Bransford goes on to state that companies such as Cyveillance "empower consumers to get the word out effectively," meaning to get their message to the people who can make a difference.

Cyveillance utilizes two proprietary technology platforms that include crawling the Internet, in addition to agent-based technology that can be targeted to online areas where relevant hot pockets of dynamic content generation are occurring. The data is aggregated and then scored to filter out the noise, leaving the relevant information for analysis. Cyveillance professional service individuals then review and polish the data before it is published in a secure area for the client to access. Included in the reports are links back to the original source (site) if the client wants to read the source materials in context.

London-based Infonic Ltd has been analyzing online discussions for corporations since 1997, and services such clients as British Airways, Bayer AG, and Levi Strauss & Co.

While technology helps locate relevant areas of interest pertaining to the client, research individuals do the heavy lifting of mining the actual discussion content by rewriting abstracts of the content and ideas found for the client. Links and references to the source are always provided, and what is stored are bookmarks to relevant notes about the information. "Most of what we offer is the understanding of and reporting of what's being said online," says Roy Lipski, managing director for Infonic.

Infonic has been advised that its process is acceptable under the Fair Use for Research precedent. The most likely items of content are downloaded, indexed, and then the downloaded materials are destroyed.

Page 3 of 4