Semantic technologies have been increasingly featured in the news in the past year, promising to provide the contextual relationships and meaning that traditional keyword search engines lack. FT Search, Inc., an independent entity within The Financial Times Group, has just launched a beta of Newssift (www.newssift.com), a new business search engine that uses semantic algorithms to provide meaning-based results. Employing facets (information categories) and Guided Navigation search technology from Endeca (www.endeca.com), Newssift lets users refine, expand, and explore. Newssift enables users to search thousands of editorially selected global business news sources, including but not limited to the Financial Times. It then offers meaning-based results that contextualize the trends, opinions, and qualitative events that shape business decisions and impact corporate reputations. It is designed to cut out the "commercial clutter" found in traditional keyword search tools."Unlike traditional search, Newssift was developed to enable users to string together a query that can provide insight into the relationship between people, organizations, geography, and business theme, which ultimately facilitates more informed business decisions," says Robin Johnson, CEO, FT Search. "There are only a few search engines that employ relationship-based or semantic algorithms, and to date there is no other that accomplishes refinement using a business point of view." Newssift is Johnson's brainchild.
John Greenleaf, chief marketing officer of FT Search, says the search tool has been in a private, invitation-only beta, with testing by some key financial people and employees within Pearson, PLC, owner of the FT Group. Greenleaf says the new tool leverages a number of key components-the current opportunity in the search market for an improved, semantic-based search experience; a handpicked, focused set of business information sources; "breakthrough technology" from Endeca; and the 100-plus year legacy of the FT brand.
Newssift looks different from the major search engines and acts differently too. If users can get past the concept of just typing one or more words into a single search box and, instead, embrace the concept of building a query in multiple search boxes using the suggestions and available refinements, they'll find a powerful discovery tool for business research. The big difference is that there are multiple boxes that populate with contextual suggestions when a user starts to type a word into the search box (see the screen example). Users can refine queries across multiple search options, such as topic, organization, place, person, and theme. Terms in each of the boxes can be further refined. Expand lets users broaden a search by choosing a term that includes a wider range of results, while Refine lets users narrow a search by choosing a more-specific term that will help them drill down into a subject. Librarians will love it-it's an information-seeking model we've tried to teach for years.
Results are relevancy ranked by default but can be sorted by date. In addition, results can be refined by timeframe, article source, and sentiment-positive, neutral, or negative. FT is reportedly not given priority in search results, but it does appear as a separate option under Newspapers in the Article Sources refinement feature. Summary abstracts are generated for the results and are actually descriptive, not just the first two sentences of an article.
Greenleaf explains that "Newssift thinks the way business people think and interacts directly with the user, offering results that are meaningful, relevant and ultimately pertinent to important metrics like stock price valuation and corporate reputation."
At this point, Newssift includes only English-language sources, though drawn from around the globe. There are no plans to add other languages in the near-term, though Greenleaf said it could be a possibility in the long-term. It includes news portals, magazines, newspapers, newswires, research sources, online news, television and radio, and expert commentary.
Greenleaf said that FT Search is working with an advisory board of advertisers, looking at testing new business models and types of advertising for the site. The company may consider other revenue possibilities in the future, including charging for some services. Greenleaf sees Newssift in a complementary position between the general keyword search engines and subscription-based business research services. Newssift is in open beta, which allows FT Search to evaluate user behavior and gather feedback on many features of the tool.
In the future, Newssift will become an increasingly customizable tool. New features and an array of widgets are in the pipeline, e.g., the ability to set up alerts and feeds, participate in Newssift forum discussions, and more. The site welcomes user feedback and suggestions.
Andrew Conry-Murray, writing at the InformationWeek blog (www.informationweek.com/blog), was impressed with the filtering, sentiment analysis, and sources. But not all his comments were rosy. "The variety of options takes some getting used to. As I fumbled around building queries and tinkering with filters, I found myself longing for the type-and-shoot simplicity of Google. And frankly, the site is ugly. Google and Apple have raised the bar for clean and attractive interfaces. Newssift still has some work to do on this front. That said, I think the site has the potential to be a very useful tool, particularly when doing deep-dive research."
Erick Schonfeld, blogging at TechCrunch (www.techcrunch.com), makes several important points about the new search tool, including the discovery component. "I am not sure I would use Newssift every day to stay on top of the latest news, but I can see it as a useful research tool when I have to really dig deep into a topic. It does better with business news than technology. Still, it is worth checking out in that it employs several subtle navigational techniques that make it more of a discovery engine than a search engine."
A number of bloggers have commented that they are intrigued by the sentiment analysis feature. This technology is supplied to Newssift by Lexalytics, Inc. (www.lexalytics.com), a company used as a partner by many search providers and services, including Endeca, FAST, Dow Jones Factiva, Thomson Reuters, and others.