Automated Service for Product Offer Creation on the E-Trading Platform with Marketplace Technology

Different information services play a major role on electronic trading platforms with marketplace technology, helping buyers and sellers to search contractors, perform feasibility analysis on potential transactions and rational selection of supply and demand parameters. Usually different filters based on the relational algebraic operations are used to contractors search and clustering. Experts-consultants are involved for decision support in more complex scenarios. However, information about supply and demand in e-trading platforms is complex, contradictory and rapidly changing, which makes it necessary to automate decision making processes.Models and algorithms for information seller's support during creation of rational product offer, focused on current generalized demand and ensures maximum of expected number of transactions are described in the presented paper. Authors propose set of models for parametric formalization of individual demand for each buyer and individual offer for each seller using linguistic variables, describing their characteristic parameters. Methodic for building membership functions of local linguistic variables (with individual parameters) of generalized demand and calculation of local matching of mentioned fuzzy values of generalized demand to individual offer. Calculation of matching generalized demand and individual offer is performed via aggregation of local matchings. As aggregation operator discrete Choquet integral over fuzzy measure is proposed. Fuzzy measure is chosen in a class of Sugeno alpha-measures. Calculated matching is interpreted as probability of transaction. For individual offer selection which maximizes expected number of transactions genetic algorithm is used. Search space is represented as parameters of seller's supply membership functions.Example of the numeric implementation of proposed service is provided.