A video data management service with context-aware recommendations
Students Name: Pisotskyi Marian Oleksiiovych
Qualification Level: magister
Speciality: Computer Systems and Networks
Institute: Institute of Computer Technologies, Automation and Metrology
Mode of Study: full
Academic Year: 2022-2023 н.р.
Language of Defence: англійська
Abstract: In the master’s qualification work, a video data management service was designed and implemented, including a context-aware recommendation algorithm. The software product is fully English-language, multi-platform and can be run both locally and on a remote server. The developed video data management service provides context-aware recommendations of video data, support for up to 5,000 simultaneous client connections to the server part of the service, and a response time of the server part of no more than 200 ms. The problem of developing a video data management service was analyzed and considered. The application of recommendation systems is considered. The concept of context-aware computing is analyzed. The software tools and technologies of the video data management service were reviewed. The choice of technologies that are the most optimal for the development of a software product is justified. The design direction of the video data management service with a context-dependent recommendation is well-chosen. The use of the multi-platform concept is substantiated. The use of client-server architecture is justified. The design stages of the video data management service are described. The structure of the video data management service is described. The software architecture of the video data management service is described. The video data management service algorithm is described. The context-aware recommendations algorithm is described. The use of similarity measures in the recommendation mechanism is considered. The user context model is considered and described, as well as options for its extension. The algorithm of the video data management service includes the algorithm of the server and client parts. The server algorithm contains the necessary sequence of operations for working with functional modules of the service. In turn, the algorithm of the client part describes the sequence of steps to provide the user with access to the functionality of the system, which is implemented by the server part. In particular, the algorithm of the client part contains the sequence of actions of the user registration/authorization process, the process of entering a login and password, as well as checking the correctness of the entered data, etc. The context-aware recommendations algorithm involves saving information about the user’s interaction with the video and information about his preferences; defining the user context; forming a profile of user preferences; forming a profile of user preferences in context; determining the similarity of a video profile to a user preference profile with or without context; determining the contextual relevance of the video; determining the overall compliance rating. The process of developing a video data management service with context-aware recommendations is described. The database scheme of the video data management service is described and considered. The class diagram of the video data management service is described and considered. The implementation of the main classes and methods of the video data management service is described and considered. The implementation of the user interface of the video data management service is considered. The diagram of interface transitions is given. It should be noted that this software product is presented in the form of a multi-page website, in other words, a web application. The client part of the video data management service is implemented in the object-oriented Java programming language using the Hibernate server framework and the XML markup language. The database is implemented using Microsoft SQL Server. The results of research, testing and debugging of the video data management service are given. As a result of load testing of the server part of the software product, it was established that it provides support for up to 5,000 simultaneous client connections, and the response time of the server part is no more than 200 ms. The economic justification of the developed software product is presented. Development costs were calculated, as well as a comparison with an analogue was made and its advantages were determined, namely: reliability, universality, speed of operation, etc. So, according to the conducted economic justification, this project solution is fully competitive. In addition, a positive economic effect in the amount of UAH 53,437.67 was obtained, therefore the development and implementation of this design solution is economically feasible. The object of the research is a video data management service. The subject of the research is the analysis of the problem of storing and viewing video data on the Internet. The purpose of the research is to develop a video data management service that will be more effective than similar ones. Keywords: video data management service, context dependence, context-aware recommendations.