Agents in E-Learning in the Business Sector

Interest in the integration of so-called intelligent or pedagogical agents in e-learning, both in business and other settings, is growing. An agent may operate within online instruction as an animated object such as an avatar, cartoon character, a disembodied voice, or, perhaps, a simple text display. The agent monitors the learner’s progress, recommends alternative activities or exercises, or makes other instructionally related recommendations. Agents may also provide useful information to instructors and instructional designers regarding student progress and needs while also suggesting possible design modifications to the instruction. This entry explains how agents function in e-learning and explores extant and potential uses of agents in the business sector.

While use of e-learning in business and higher education differs in regard to context, purposes, learner characteristics, and situational factors, appropriate utilization of agents in both contexts can increase the responsiveness, content, quality, and relevance of online instruction, leading to improved efficiency and effectiveness. Nonbusiness related work on use of agents should be applicable in business.

Dawn G. Gregg suggests that agents have high potential for supporting learning. Such support entails providing instruction based upon the learning needs of a given student. Gregg further notes that agents are programs that can function to shroud complexity, complete learning tasks, provide instruction, support multilearner collaboration, or track learning processes and events.

Beverly Woolf and Chris Eliot propose an expansive learning grid entailing use of multiple computing resources and learning institutions. This model incorporates six types of agents:

Woolf and Eliot see their learning grid as a means of customizing instructional resources on an individualized basis so as to improve the quality of instruction based upon a student’s instructional history. The following are examples of each agent function Woolf and Eliot suggest:

Monitoring of performance: This general function might best serve as both a data collection mechanism and as a feedback function to the learner. An example would be simple monitoring of the completeness of student input and alerting the student when information is missing, similar to the alert one receives when filling in credit card information and leaving out the security code.

Suggesting resources: Based upon instructional goals and purposes, this function is self-evident and can include providing additional instruction or readings to improve poor learning.

Suggesting resources based upon metadata: This function clearly requires ongoing collection of performance and instructional progress data from multiple users. An example would be providing the user with resources drawn from previous learners’ performance, such as suggesting a resource on how to conduct a Boolean-based search in a library database because metadata suggest that many end users are unsuccessful in their topical searches.

Suggesting lesson redesign improvements: An example here would be to recommend revising instruction for given learning objectives based upon learners’ poor performance on one or more related test items.

Combining and reporting collected performance data with student evaluative data: This function and the next imply the need for direct instructor involvement; for example, the instructor analyzes course evaluations by students and test scores by objective then revises the instruction as appropriate.

Coordinating instructional progress data with instructor advice to suggest required additional instruction: This function, similar to the previous, might entail analysis of software generated practice or test performance data and related problem areas for students with individualized or whole-group directed advice to learners from the instructor. Such instructor-based advice might be grounded in the instructor’s experiences in teaching the given content, intuition, and knowledge of problematic aspects of the instruction for learners over time.

...