Optimizing Trading Recommendations in Portfolio Trading: A Bilateral Matching Theory Approach

The development of information systems has greatly changed our way of life, and many group activities, including communication and trading, can be easily carried out online. Taking transactions as an example, including transactions of various items, online trading platforms provide a trading market, which can greatly improve trading efficiency and increase transaction volume. During trading, the market maker platform will assist customers in providing trading guidance services and facilitating a certain number of transactions when it is not possible to trade all. Examples of portfolio trading mainly include trading in the second-hand goods market, stock portfolio trading, and some small markets aimed at completing specific transactions. The transaction recommendation system is a user recommendation system based on a portfolio trading market matching mechanism algorithm. This trading recommendation system takes user information as input. This paper constructs a mathematical model of the market based on bilateral matching theory, and also visualizes it into a weighted bipartite graph. The parameters are obtained by solving the model based on the interior-point method and revised simplex method. The system feeds back the algorithm calculation results to users in the form of recommendation indices. The transaction recommendation system can be applied to software, web pages, and other trading platforms that can utilize backend computing power and have the ability to collect user information. The transaction recommendation system directly serves users.