The Design And Implementation Of Advertising System For Large-scale Video Websites Based On Cloud Computing Platform

Internet advertisement putting system is generated in the fast development of online advertisement market and Internet technology, aim ing to help advertisement owners and Internet enterprises, including online video service providers, to realize professional and accurate online advertising, advertising management, delivering statistics and other management functions. As the leading video sharing website in China, 56.com daily video playing amount exceeds ten million times, the company’s objective of building online advertisement putting management system is to make deep integration with sales team and form diversified sales plans through strict control of putting process and data, depending technical basis of advertisement putting system and under the premise of ensuring accurate advertisement putting, it aims to ensure efficient use of advertisement resources through more fine management of company’s media planning and product sales persons and timely opti mization of advertisement putting plans.The thesis explores computational advertising, process management principle, similarity recommendation algorithm, elastic computing, massive storage technology and relevant principles and technology, uses cloud calc ulation platform to build an online advertisement putting system of online advertisement intelligent putting, putting process management and putting data digging, which supports daily one million visits, thus enabling users to effectively control visit of various advertisements in different terminals, different periods and different places, it can provides different putting plans for advertisement owners of different types, and greatly increase exposure frequency of the brand in objective crowds.During the development process of the system, it contains deployment of system’s hardware groups, strategy of advertisement putting control, development of putting management system software, analysis and forecast of massive data, mainly realizes following several functions, it builds putting sub-system and data collection subsystem on the rear end of advertisement, realizes advertisement putting control and data collection; designs and develops process management sub-system which is used for reservation, management and advertisement resources management; designs and develops advertisement contract management, ordering management sub-system, realizes accurate putting and similarity recommendation for users’ images; designs and develops putting data monitoring and effect feedback sub-system; designs and develops data counting sub-system of advertisement system.Through different levels of tests for each subsystem, the advertising system to meet the proposed 56.com functional and non-functional requirements, and on this basis to achieve based on the Pearson correlation coefficient, Cosine similarity, Tanimoto coefficient use groups wisdom collaborative filtering recommendation algorithm advertisement.