Archives of computer-mediated communication (CMC) could be valuable organizational resources. This paper describes an automatic approach that facilitates users' integrated understanding of discussion and behavior of CMC participants. In the presented parallel learning framework the global population is divided into some subpopulations, each assigned to a distinct processor. Each subpopulation consists of the same class fuzzy rules. These rules evolve independently in the proposed parallel manner. Second we validated the approach through the development and evaluation of a prototype, the Communication-Garden architecture and technology to fulfill the archive users', we propose three-tier architecture. It includes information representation, categorization, and visualization. The representation level applies automatic indexing techniques to extract semantics to represent message content, while the categorization level employs different analysis technologies to extract patterns and salient structures from the archived information.