The Facet Advantage
Information retrieval system design vs. user skill

In the world of information research there is a fundamental question: Does system design or the user’s research skills produce quality search results? The answer has always been, “both,” but, with the implementation of some new search methods gaining popularity, the balance is shifting, and technologies are making greater contributions to the production of high quality results.

Over the years, information retrieval systems designers were faced with two major problems. One, they had to determine how to make the systems better and two, they needed to make the users of these systems more efficient at searching the data. Since it was a nearly impossible task to train every possible user in research skills, system designers have focused on system improvements geared toward producing high quality search results with minimum user effort.

The first notable change came with the introduction of searchable subject headings clusters of MESH in the CANSEARCH databases back in 1987. The idea driving the database was to keep users from needing complex command based Boolean searching skills and enabled them to browse clusters of documents related by subject. This was one of the most revolutionary conceptual changes in system design. This paradigm shift was further exploited by the development of the web environment.

Information vendors quickly adopted the advances in web based applications and new user focused information retrieval systems were introduced. These systems provided indexed search settings and automatic truncation along with pre-defined web-based search forms. These advances gave the user confidence in their abilities to perform searches and get results that provided relevant information. Hyper linking was also utilized in order to present results and related terms in a way that made relevance judgments easier.

However, these advances in systems design still did not address the user side of the equation – Users remain reluctant to utilize these features and flexibility, preferring to plug keywords in a search box and hit ‘go.’ We jokingly refer to this as “The Google Mentality.” Every search was left to chance and the user would rely on the first dozen results to provide the ‘answer.’ Understandably, more experienced researchers are left frustrated and dissatisfied with the results – quickly finding them irrelevant to their query.

But looking at recent advances in search technologies, we can see new approaches to this problem and among the most exciting are search results presented in facets.

A facet classification does not assign fixed slots to subjects in sequence. Instead it uses clearly defined, mutually exclusive, and collectively exhaustive aspects, properties, or characteristics of a class or specific subject. Such aspects, properties, or characteristics are called facets. In almost every commercial site today, from Amazon.com to Market Research, there is an option to browse by subject. This means a user is presented with pre-organized data according to general subjects and they can search within these specified fields. This eliminates the gap between the user’s conception of terms and the system’s pre-assigned terms and enables more precise queries of the data.

The continuous improvements in design leave much less room for error and even an inexperienced user can obtain relevant results. So we can see that through smart design and implementation, both questions are answered. If we make the tools more intuitive and powerful, the end-user automatically becomes a more effective searcher.

BACK TO TOP

Autostemming is Now Available on Expert Search!

One of the enhancements implemented in the new release of Engineering Village 2 is the addition of Autostemming functionality to Expert search. The Autostemming function allows the searcher to control the key word search by selecting the system to Autostem the searchable key words.

By adding this functionality to the Expert search interface, greater flexibility is achieved by allowing the user to choose which of the key words in the search will be stemmed. Perhaps the user wants to use a combination of autostemmed and non-autostemmed terms.

As seen in the example below, the phrase “micro computer” was put into the system, while the word ‘Micro’ was stemmed and “Computer” was not.

Click on screen to enlarge

Click on screen to enlarge

For more information about Autostemming see our previously publishes article at: http://www.ei.org/eiupdate/archive/
issue6/03_librarians_corner/index.html#ei_p3_01

To see how your clients can get the most out of their research, please contact us at Customer Support.

BACK TO TOP

Yesterday’s Research Powering Today’s Innovation

In order to provide the tools needed for staying ahead in today’s technology race, researchers are looking to the past and finding success. This trend can be seen in the numerous products currently on the market employing technology that was first used over 100 years ago. The technological “winners” are those clever enough to use yesterday’s discovery to drive today’s innovation.

As a student of Hertz (as in Mega- and Giga-), Guglielmo Marconi perfected the use of radio waves for communications in the shipping industry. The result was the wireless telegraph and the ‘walkie-talkie’ style of radio communication. But this method of using radio waves fell out of favor very early on when the use of frequencies allowed multiple transmissions without interference. The assignment of distinct frequencies to different tasks then gave rise to modern commercial radio, television, and satellite communications – each representing a tremendous commercial success.

Today, many of us live in a world of wireless speakers, the wireless computer mouse, wireless headphones and the wireless home network. Wireless features may be the cutting edge of personal electronics, but the technology driving it dates to Marconi’s early work. Bluetooth is an application of wireless technology in the personal electronics markets that is steadily growing in popularity. It solves the problem of wireless transmission through crowded frequencies by transmitting data over unused bandwidth in the radio wave spectrum at very low power levels. It’s the modern equivalent of Marconi’s wireless telegraph adapted to meet today’s wireless networking needs.

Will the “next big thing” be based upon a technology invented decades ago? Proponents of UWB (ultra-wide band) communications think so! UWB is being positioned to replace Bluetooth in the area of wireless communications and is based upon research on microwaves first conducted in the 1960’s. So while engineering innovation may be cutting edge, there remain strong ties to work done long ago. Today’s successful researchers may find inspiration in the work of their predecessors – and those who continue to overlook older works do so at their own peril.

BACK TO TOP