Scenario for Generic Project

A Student asks a question containing certain keyword. This is a short message (message id equal to 8). The message is broadcast to all and received by Intelligent Agent and Certification Agent.

Intelligent agent sends short message to Consultant to find the answer. Intelligent agent can send short message to Retrieval Agent to narrow down answers. Intelligent agent can also poll the students to find a majority answer for the question.

Retrieval Agent uses keyword to retrieve relevant text and e-mail to student who asked the question.

Certification Agent verifies every question posted by student is eventually answered by someone: a human consultant, a Retrieval Agent, the instructor, or a fellow student.

Consultant Agent is the front end for the human consultant, who can define a profile, e.g., to (a) always answer the question, (b) sometimes answer the question, or (c) never answer the question.

Note: We can add message sub-types in the body of the message so that no new message types need be introduced.

Project Groups

G1: intelligent agent
G2: retrieval agent
G3: intelligent instructor agent
G4: intelligent chat agent
G5: retrieval agent
G6: consultant agent
G7: retrieval agent
G8: retrieval agent
G9: intelligent agent


Message Type: Short Message ID=008
Local Syntax for Message SubType:
  KEYWORDSET:ski,tree    (set keywords)
  LOCATION:pittsburgh    (set location)
  TALLY           (tally students when session is over)
  ?             (a question mark anywhere indicates a query)