| |
The paper describes the new information system architecture model named The Librarian Model. The model is developed for the automated information systems. The article describes an architecture design model for process automation and the organization of automation design architecture, a model called the Librarian model for designing an automated system architecture.
Development of an architecture design model for automated information systems, taking into account the load in the form of an incoming stream of service requests, in particular, for writing to the database. Investigate the model of approximation of the exponential distribution law of events of signal arrival to the Poisson stationary process of event flows.
Methods: Methods of mathematical and static modeling were used to study the behavior of the occurrence of events, according to the exponential distribution law, and a model was developed for the transition to the Poisson event service time flow. A model of the architecture for processing multiple requests based on synchronization based on the messages of incoming requests for writing to the database has been developed.
Results: The criterion of waiting time for servicing requests was studied, by which the degree of approximation of the flow of events to the Poisson flow was estimated, the use of which can help to adequately evaluate the planning of system resources and the design of the model of the required architecture. Based on the studied behavior of the occurrence of events in automated information systems, the Librarian's model is proposed, which takes into account and adequately distributes the dynamic load in the form of a Poisson flow of events.
Conclusions: Further research in this direction may be related to the practical use of the criterion for evaluating the properties of event flows and the estimation of errors, which are caused by replacing non-Poisson flows with Poisson ones when modeling systems, as well as using adequate architectures for automated information systems when designing, taking into account the evaluation of computational time to service many user requests, as well as team exchange with third-party platforms.
Keywords: architecture, information systems, automatization systems, recommendation system, recommendations, collaborative filtering, element-based, information system types-based, content-based filtering, knowledge-based filtering, hybrid recommender systems.
|
LEGAL
INFORMATION:
