Information Product-oriented Data Resource Management Methods For Social Media Platforms

It only costs about ten years for social media to grow from simple networking services to full-scale web and mobile applications that are deeply engaged in everyday life.However,most social media companies face a lot of problems such as low stickiness,weak innovation and lack of competitiveness.The main cause is that they are still following the traditional technology-centric stretagy,which brings a highly homogenization trend among different platforms in terms of functioning,management modes,user interfaces and even content.Given this background,the big data resource on social media platform is a new source of opportunity if it can be collected,used and transferred into high-quality customer services.Standing on the view of social media firms and targeting at producing high-quality information products,this thesis uses an information-product-oriented perspective and follows design science approach,explores the methods on social media data resource management.The work is organized as follows.Firstly,it defines and charaterises social media data resource and social media data product,and proposes an “information product as a service” mode for social media platforms.It then compares the three pespectives: technology-centric,business analytics-centric and information product-centric.A framework for strategiccally managing social media data resource as a product is presented.Connotations and managerial approaches of the four main principles—needs,supplies,production and quality are explored.Details of principles and their implementation are further discussed using an online nagetive word-of mouth handling case.Secondly,aiming at building a systematic conceptual model of social media data resources,this thesis develops a standardized “Actor-Action-Object” framework that describes users' various online activities,which are the of corresponding data resources,and proposes activity primitives,activity components,and forming operators.The conceptual data model has great implications for understanding the data resources' ultimate origins,taxonomy,distribution,and characteristics,and help platforms draw up adapted data resource management strategies to achieve more efficient resource development and value transformation.This conceptual model is superior to current models in integrity,ease for use,understandability,and expandability.Thirdly,by introducing hypergraph theory,this thesis proposes an undirected hypergraph model for social media data resources.It presents the modeling process,describes the advantages of model in describing complex entities-relations,modeling flexibility,expandability and supporting data processing,and then proposes several important data proceesing schemas based on the hypergraph model.It focuses on the dynamic feature of the model and studies assimilation-based and accommodation-based evolutionary mechanism.Fourthly,it analyzes the decision problem on which production line to proceed in social media data product manufacture.By comparing the characteristics of production lines of information product and physical product,this thesis interprets the necessity of an inclusion of all three decision elements—supply,demand,and process,and considers quality as of huge importance in decision process.The quality aggregation function and quality transfer function are proposed to explain how data quality are changed during the process of product assembling and manufacturing,respectively.Based on directed hypergraph theory,it develops a directed hypergraph model for manufacturing system and proposes several algorithms that suppot the production line decision.An experiment is conducted to evaluate the validity of the model.Finally,the contents and conclusions of this thesis is summarized,and the perspective for future work is presented.