INTELLIGENT AGENTS

There are many online information retrieval and data extraction tools available today. Although these tools are powerful in locating matching terms and phrases, they are considered passive systems. Intelligent Agents (see Watson, 1997; Bigus & Bigus, 1998) may prove to be the needed instrument in transforming these passive search and retrieval systems into active, personal user assistants. The combination of effective information retrieval techniques and intelligent agents continues to show promising results in improving the performance of the information that is being extracted from the WWW for users.

Agents are computer programs that can assist the user with computer applications. Intelligent Agents (i-agents or IAs) are computer programs that assist the user with their tasks. I-agents may be on the Internet, or they can be on mobile wireless architectures. In the context of this research, however, the tasks that we are primarily concerned with include reading, filtering and sorting, and maintaining information.

Agents can employ several techniques. Agents are created to act on behalf of its user(s) in carrying out difficult and often time-consuming tasks (see Jensen, 2002; Watson, 1997; Bigus & Bigus, 1998). Most agents today employ some type of artificial intelligence technique to assist the users with their computer- related tasks, such as reading e-mail (see Watson, 1997; Bigus & Bigus, 1998), maintaining a calendar, and filtering information. Some agents can be trained to learn through examples in order to improve the performance of the tasks they are given (see Watson, 1997; Bigus & Bigus, 1998).

There are also several ways that agents can be trained to better understand user preferences by using computational intelligence techniques, such as using evolutionary computing systems, neural networks, adaptive fuzzy logic and expert systems, etc. The combination of search and retrieval engines, the agent, the user preference, and the information retrieval algorithm can provide the users with the confidence and trust they require in agents. A modified version of this approach is used throughout this research for intelligent information retrieval from the WWW.

The user who is seeking information from the WWW is an agent. The user agent may teach the i-agent by example or by employing a set of criteria for the i-agent to follow. Some i-agents have certain knowledge ( expressed as rules) embedded in them to improve their filtering and sorting performance. For an agent to be considered intelligent, it should be able to sense and act autonomously in its environment. To some degree, i-agents are designed to be adaptive to their environments and to the changes in their environments (see Jensen, 2002; Watson, 1997; Bigus & Bigus, 1998).

This research considers i-agents for transforming the passive search and retrieval engines into more active, personal user assistants. By playing this role, i-agents can be considered to be collaborative with existing search engines as a more effective information retrieval and filtering technique in support of user needs.