If Search is So Great, Why Are We Still Searching?
In spite of the promises made by search engines and portal vendors, employees are still wasting time reinventing the wheel when they start a new project, taking too long to track down files.
The problem is that, while portals play an important role in leveraging existing structured data held in legacy systems, they are not as good at accessing unstructured data—the "stuff" that makes up 85% of the information in an organization.
Search engines were developed to try to make sense of unstructured data. However, these engines are not perfect—anyone who has looked for information on the Web using traditional search engines knows how frustrating and time-consuming it can be.
This is because users are really searching for subjects, not keywords. For example, users looking for "Oracle" might be interested in either the "Oracle of Delphi" or "Oracle," the company. They might be interested in not just articles that contain the word "Oracle" in every sentence, but in related companies, employees, products, news articles and technologies.
But typical keyword searches don't provide these paths to data discovery. They only provide pages and pages of search results that weakly reflect a user's queries.
A complete solution is the key
The ability to automatically create structure from unstructured data sources for use in information retrieval tasks represents the most promising approach to solving the growing problem of information overload and inaccessibility.
A successful solution should be advanced, comprehensive and designed specifically for unstructured data discovery. Components should include multilingual language analysis, search, data classification and taxonomy support, intelligent summarization, entity extraction and visualization technologies.
Such complete solutions reveal information and relationships previously hidden in unstructured documents, enabling users to both search (for information that they know exists) and browse (to uncover new information).
The typical user is familiar with standard search—a box into which a one- or two-word query is entered. It is, however, what happens after the user hits the "search" button that is the key to a successful search.
For example, a search for "table" when done solely with keywords would result in a random page after page mix of documents (ranked by popularity) on pieces of furniture, charts used for data presentation, and the postponing action a committee might take.
However, by adding language analysis, entity extraction and classification, smarter searching is enabled. For example, applications and searchers can benefit from knowing that in one context "Georgia" is a woman's name, in another "Georgia" is the name of a U.S. state, and in another that "Georgia" is a country in the Caucasus on the Black Sea.
For more rapid digestion and navigation, search results should also display intelligently created document summaries (20% more accurate than typical search summaries), and be able to display the returned information in either a standard text tree or in a visual tree (proven to be 65% faster than accessing text information).In short, without a complete suite of tools working in harmony, search scenarios result in wasted efforts and frustration.
Complete, powerful, simple methods for finding information
Inxight gives customers a complete and scalable solution that addresses all of the essential elements of unstructured data search and information discovery. Inxight's solutions are based on the industry's most powerful natural language processing of text data in over 26 languages and more than 70 text file formats.
Inxight's software reduces by over 60% the time users spend finding vital information from any text data set, regardless of origin, location, language or complexity—giving knowledge workers more visibility and reducing wasted time, adding millions of dollars to the bottom line.
Government agencies are using Inxight SmartDiscovery to uncover terrorist networks and sort through millions of agent reports daily. Publishers are using SmartDiscovery to automatically categorize their intellectual property for faster, more relevant distribution. And enterprises of all flavors are using SmartDiscovery for faster drug development, for locating information buried in emails to comply with Sarbanes-Oxley and for improving productivity and reducing frustration.
The companies and organizations that are investing in a complete search and information discovery solution today are going to be the companies and organizations that lead the economic recovery. Those that don't will still be "searching" for productivity and growth for years to come.
Inxight provides the fastest route from search to discovery. Using automated information extraction, categorization, search and visualization software, users find the precise, relevant information they need—boosting productivity, accelerating time-to-market and leveraging their past and future information investments