Research On Fair Use System Of Copyright In Data Mining

The computer processing process of extracting unknown and potentially relevant information from data is called data mining,which is based on the principle of copying a large amount of original text,extracting the data from it,and reassembling the process to obtain new information that is potentially unknown.With the advent of the era of artificial intelligence,the research approach to problem solving has changed from a hypothesis-based approach to a discovery-based approach,so data mining is undoubtedly the best tool for researchers to deal with massive amounts of information.Because of its advantages of reducing data processing costs and improving analysis efficiency,data mining has been widely used in several fields,and it is not only an important pillar of the data industry,but also of great economic utility.However,there is a certain risk of infringement involved in terms of the processing model of data mining.The implementation of data mining often requires the copying of materials,as the data it mines includes both copyrighted and non-copyrighted materials.When only the target text is processed separately and only a small amount of data is copied each time and no fixed copies are grabbed,data mining generally does not constitute an act of copying under copyright law,but when a large number of copies of the target text are made,data mining is caught in the dilemma of copyright infringement.That is to say,the process of data mining to extract information will involve a large number of copyrighted works,which will inevitably infringe the copying rights of copyright owners when authorization is not obtained or the exemption conditions are not met,while the process of data mining to identify and transcode the extracted data may also infringe the deduction rights and information network dissemination rights of copyright owners.Since it is obvious that it is too costly and not easy to seek licenses one by one,and may even cause restrictions on the application of data mining,providing exemptions for data mining through the fair use system is undoubtedly the best choice.At present,China’s fair use system has not yet made specific arrangements for data mining,which will easily lead to different judgments in disputes due to the inconsistent scale of adjudication.In order to alleviate the legal crisis faced by data mining and improve the current fair use system,it is necessary to explore and analyze the above issues from the theory and practice of China’s copyright law to determine whether data mining is established as an infringement in China’s Copyright Law and whether there are defenses,and to propose amendments to the Fair Use System.Specifically,there are already countries in East Asia that have taken action to introduce specific fair use exceptions for data mining,or by having a flexible fair use factor test that allows local courts to cover such cases.Given China’s social,cultural and political environment,the design of the data mining fair use system should follow a semi-open approach,while the three-step test for determination needs to be clarified in practice,supplemented by the four-factor analysis method of transformative use judgment,to further promote the use of the fair use system and allow data mining to maximize its benefits.