Relevant Technical Research On Trading Partner Intelligence Found Based On Ontological Multi-agent System

With the rapid development of information technology and the growing popularity of computers, a variety of Internet applications emergs. E-commerce is one of the most prominent applications, and is used widely driven by Internet technologies,which not only broke the time and geographical constraints, but also completely changed people's lifestyle, greatly improving the ease of market transactions and flexibility. Out faster creating a rapid development of e-commerce, it also brings many problems, such as low efficiency brought about by information overload, high transaction costs, unity of the system features, high difficulty to find trade partners between buyers and sellers and so on. Therefore, in order to more efficiently achieve the discovery of trading partners, this article Multi-Agent technology and ontology technology into the environment of e-commerce transactions, First, with the Agent has the autonomy, intelligence, distribution, coordination, self-organizingmany characteristics of ability, learning ability and reasoning ability, in some cases simulating human behavior, and thus to solve practical problems in e-commerce, intelligent business processes, improve efficiency, reduce transaction costs. Secondly, ontology introduction of e-business environment, products or services to buyers and sellers in e-commerce environment ontology, said to help buyers and sellers to quickly match the right products or services and consultations, the final completion of the transactionA tree similarity algorithm for match-making of agents in e-business environments was presented. Product/service descriptions of seller and buyer agents were represented as node-labelled, arc-labelled, arc-weighted trees. A similarity algorithm for such trees was developed as the basis for semantic match-making in a virtual marketplace. The trees were exchanged using a serialization in Object-Oriented RuleML. A corresponding Relfun representation was employed to implement the similarity algorithm as a parameterised, recursive functional-logic program. Our tree similarity algorithm, as part of semantic match-making, computed the similarity of subtrees in a recursive way. It could be parameterized by different functions to adjust the similarity of these subtrees. This gave agents the option to tailor their similarity measure for trees based on their subjective preferences. The appendix gave the full definition of the algorithm in Relfun. This executable functional-logic specification has proved to be a flexible test bed for our experiments. Results from our experiments were found to be meaningful for e-business/e-learning environments. The algorithm can also be applied in other environments wherein weighted trees are used.This multi-agent system could be further improved by using clustering techniques. Based on the similarity between every pair of trees, the future multi-agent system should have the ability of clustering agents according to selected clustering algorithms based on our tree similarity measure.