چكيده به لاتين
Nowadays, users have access to a large amount of information due to the expansion of the Internet and the development of various media platforms. Depending on the growth and variety of data, it's harder to find items that fit the needs of users every day. In such a situation, the use of Recommender systems can be very helpful. The goal of the Recommender systems is to select and provide items or services for users that have the most relevance to their needs.
Traditional recommender systems offer their suggestions regardless of the context and environment of the user, although choices and user’s needs are affected by circumstances such as time, place, etc. the context in which the user interacts with the system has a significant impact on the type of behavior and selection. As a result, systems that consider the context of the user can provide suggestions that will increase the user's satisfaction from the system, and clearly illustrates the importance of research on context-based recommendation systems.
One of the challenges faced by context-aware recommender systems is domain dependency. This dependency means that the application domain is very influential in the selection of context. For example, time may be considered as the most important context in one domain, but in other domains, it's not just context. Setting up and developing context-aware recommender systems, based on the domain-specific nature of the context, involves financial and time-based costs, which in the current era, would delay the decision-making and proposing irreversible consequences every second. In this research, we tried to present a comprehensive method that considers the properties of the context and, as far as possible, free of domain. We then implemented our general framework on a standard dataset, the results of which represent the proper function of the framework provided.
