To see a new car model or maybe find a great recipe, all you need to do is enter search words into a public search engine like Google, and something useful will come back in no time. But what if you are a serious researcher faced with issues like developing and launching new drugs or selecting the location to drill an oil well? Millions of dollars are riding on the decision, and you need confidence that your search as has been thorough. You can ask Google and get millions of items in a fraction of a second. But are you willing to rely on a first page result, or look at millions other results to be sure? Serving stakeholders interests and your own career are at stake.

Google and other public search engines are really good when all you want is a search of the chaotic surface web. Experts suggest this covers less than 10% of all the data available on the web. This is because Google is designed to behave like it’s still in high school and it only asks the popular kids to dance. However, professional researchers have graduated to serious work that increases stakeholder by value finding relevant but not necessarily popular information. A deep web search can help you make better decisions through federated search technology.

If your organization is knowledge driven, you probably have access to a number of subscription data bases, structured and unstructured enterprise text collections and other specialized sources that are part of the more than 90% of the web not visible to public search engines like Google. But without federated search you spend considerable time separately logging in to each engine to generate many lists of unconnected results. In addition, you need a lot review time to cull and sort the information to help make some very important decisions.

Just like Google, a federated search is conducted through a single search page. Federated search then simultaneously retrieves matching documents from all sources and therefore eliminates much of the time you would otherwise spend searching one source at a time. It then evaluates “competing” sources and generates a single result list ranked by relevancy. This deep-Web search can immensely help you, as a researcher, to focus your time on more fully understanding opportunities and risks, and improving the quality of decisions. While relatively new technology, there are several federated search capabilities available on the market today. They all deliver the time saving realized by using one search page to access multiple sources at one time.

But that’s only a small part of the benefits available to you from federated search. You should also be concerned about the application’s ability to “listen” to the specific question you are asking, and that it can further tailor results to meet your needs.

In a quest for federated search, you should be confident that the relevancy of federated search results are improved based on the specific search terms you enter, and by weighting search term usage in titles and summary text. This can be done in several ways. First, the length of the title containing the terms provides insight into the degree of focus on the research subject. Second, the actual location of the terms in the title itself suggests if the author is using them as the primary subject or just a modifier of the primary subject. A third indicator is the relative usage of search terms in the title and summary description. The result is that you feel like you are being listened as you judge the quality of returned results. Armed with a powerful result list from federated search you, the researcher, are well prepared to perform your analysis and make recommendations and decisions.

If you are part of a leading research organization, you understand that truly effective research is a process, not just a project, and capabilities such as search result refinement, reuse and collaboration significantly leverage your important research activities. Subsequent articles will explore more deeply the importance of result relevance and how can be achieved, and the impact of effect workflow for managing results.

Lance Danbe is vice president of sales and marketing at Deep Web Technologies in Santa Fe, New Mexico. To find out more about how their groundbreaking capabilities can improve your research results, and to take a federated search test drive, please visit http://www.deepwebtech.com/index.php.