Library research is facing unprecedented challenges and opportunities from big data. In recent years,scholars have been involved in the research,but the literature on the library subject in the nationwide big data survey and analysis has not been found. This paper carries out the survey with the help of the Baidu index platform and uses the visualization technology and the gray theory to analyze the survey results. This survey found:from January 1,2011 to May 20,2016,the index to the word “library” was rapidly increased from 1435 to 1859 points,a rising rate of 30%,and the index average was 1647 points. Men accounted for 73%,and women for 27%. 20-39 years old young people accounted for 75%. The crowd number concerned about the library gradually decreased from the southeast coast to the northwest inland,corresponding to its regional economic development degree. They focused on the library's inquiry,function,document downloading,open hours and borrowing mode,and also wanted to know OA(Open Access)resources,National Library of China,electronic library,etc. Media attention to the library is not high and there is a correlation between the media index and news events. This survey also found:there are deficiencies in the literature download and open service of the digital library. Although scholars and university libraries are exploring the construction of OA resources,its development trend is not clear enough. The methods and strategies at present cannot fundamentally solve the problems above,so it is necessary to put forward new ideas,methods and strategies. This author thinks that the first leap is from the paper library to the digital library,and the second one is to the big data library,which is a kind of active precision delivery system based on data mining and knowledge discovery. Any person at any time and any place can get its ubiquitous services from an Internet access device. It accepts a request from an Internet agent and provides a service to a network anonymous reader through the agent. The result the reader gets is a “knowledge module” form which may be the text,video,audio,experimental process,network connectivity,software systems and/or abstract data format. Because of the huge amount of contents,the operation of its collection and renovation needs to be automated. Raw data from different sources and different formats will be integrated and unified in order to be shareable and open,but all the contents are dispersed in a mixed cloud and the big data library system also runs in the cloud. The support system of the big-data-base is Big Table or H Base. It may be a resource polymerization of all the libraries in a field or a nation,and may even be the convergence of all the libraries in the world. Looking further into the future,a library should not wait for readers to consult the literature but not deliver accurately the desired content to users；instead it should actively participate in the global brain program and use its own big data inventory advantage to provide adequate knowledge resources for the brain reasoning machine. The Baidu index gives only seven visual results made from China's domestic data,and the next step of our research is to use the global data. 8 figs. 2 tabs. 47 refs.