یک مدل ادغام اطلاعاتی از اتحادیه های نوآوری بر اساس شبکه بیزی An Information Fusion Model of Innovation Alliances Based on the Bayesian Network

1 Introduction An innovation alliance is a union that consists of enterprises, universities, research institutes, and other institutions. These parties usually have a specific purpose and some common attributes. When information is transferred and applied in the innovation alliance, the process tends to be complicated, syncretic, and marked with other characteristics. Acquisition, classification, synthesis of the information, and decision-making require much work[1, 2]. Information fusion technology is an important tool in the field of knowledge management that can effectively organize information from different knowledge sources and form new and more comprehensive knowledge[3]. The process of information fusion in innovation alliances can collect, transfer, fusion, and apply information from alliance members and external information sources. As a result, sharing, mobility, and integration of knowledge are promoted, and collaborative innovation between each innovation subject is implemented[4] . Information fusion technology is a series of processes that enables the complete use of all types of environment and object information obtained from a plurality of sensors, analysis, and reorganization of information. Such technology allows the fusion of complete and effective information to provide support for decisionmaking[5–7]. In the 1970s, research institutions in the United States attempted to use the information fusion technology for multiple continuous fusion of sonar signals to determine the position of enemy submarines. This military application is the initial source of information fusion technology; thereafter, early forms of information fusion were mainly intended for data fusion. In the 1980s, the United States set up a data fusion technical expert group and formally proposed the concept of data fusion[8]. In 1997, the same country established the Institute of Information Fusion and organized a number of special projects to study data fusion technology. With recent rapid development in computer and network communication technology as well as the growing need for military applications, information fusion technology has seen remarkable achievements[9], and its application areas have gradually expanded from the initial military field to other areas, such as intelligent robotics, image processing and analysis, earth science, agricultural applications, weather forecasting, modern manufacturing industries, and economic management[10]. However, the modeling mechanisms and application conditions of these technologies differ, and each method presents some limitations for certain prediction problems in various applications. For example, estimates may not be accurate enough or the accuracies obtained may not be high enough. Gaining complete and effective information is key to decision making, and how to integrate various methods and overcome the application limits of application and accuracy of each method is an important area of research with both theoretical and practical significance[11] .