Use of "natural language" a step forward for search engine users
Staff -- Manufacturing Business Technology, 10/1/2003 12:00:00 AM
For years, people have been saying that search engines don't work. The emergence of Google—with its innovative approach to searches—as one of the primary gateways to the Internet may be an indication that has changed.
Another possible indicator is that a Web site knowledge base that is said to serve half of the world's design engineers at least once a month recently incorporated search technology that lets those engineers ask concept-driven questions in natural language and get results that are said to be considerably more relevant than simple key-word searches.
National Semiconductor has deployed iPhrase One Step for self-service search and navigation across its Knowledge Base (http://knowledgebase.national.com) and Solutions.National.com Web sites.
Knowledge Base includes thousands of integrated circuits, and data on more than 10,000 products, 2,000 data sheets, 400 application notes, white papers, and other information.
Solutions.National is said to be the industry's most comprehensive suite of design tools, interactive product selectors, and online information. It features 30 color-coded block diagrams organized for specific application in eight industries: automotive, broadband communications, displays, industrial, medical systems, personal/consumer electronics, power, and wireless communications.
Cambridge, Mass.-based iPhrase Technologies says its technology eliminates the most common frustration of users of search engines: wading through long lists of irrelevant results. One Step also supports a wide range of user expression, and automatically corrects misspellings, word order, synonyms, and abbreviations.
"National Semiconductor's Web site serves half of the world's design engineers at least once a month," says Andre Pino, a senior VP with iPhrase. "Its use of iPhrase One Step is a powerful example of how well One Step understands and delivers exact results against a wide range of structured and unstructured data."
