Based Smart Trading Recommendation System Design And Realization

With the development of the Internet and the e-commerce, traditional barter is facing a new turning point in the e-commerce environment. E- Barter is a modern form of barter which combines the barter with the e-commerce. E-barter acting as a cash supplement can help dealing with personal unused items, overstocking, financial issues and other problems.Traditional e-barter trading platform need brokers or intermediary help to search opportunities between the parties involved in the barter transactions who has difficult to find supply or demand in the barter market and reach deals. There are some drawbacks when barter brokers are involved in the e-barter system as follow:(1) The personal capacity of the barter broker is almost the most important point in the barter transactions, which means that costs in barter transactions has to been increased a lot with the labor costs. (2) Too much barter brokers and the lack of information sharing bring about that many potential trading opportunities cannot be found and the user interests in the transaction are finally decreased. (3) The barter transactions which are over-reliance on the ability and the experience of brokers need a system with which products can be automatically and systematically matched.On the research of the popular electronic barter trading platforms, "Directed graphs model of supply and demand", a new transaction data mining model based on directed graphs, has been proposed. The model will construct a directed graph with the transaction data and trading needs, mine this graph with directed graph related algorithms and eventually achieve automatic matching of supply and demand in an e-barter trading platform.In this paper, an intelligent trading recommendation system based on directed graph model of supply and demand has been proposed. This system transfers all the supply and demand requirements in the database into directed graphs. When there is an actual demand happens, the intelligent trading recommendation engine in the system will calculate all possible trading plans and recommend them to the user. What the users need to do is upload what things they can supply and what they want and pick the most conducive trading plan according to their personal preferences. At the same time, users may find some potentially transactions with the ancillary data offered by the recommendation engine. Experiments show that the recommendation system can improve the success rate of transactions and transaction efficiency in the e-barter trading platform.