To put it simply, literature is all the carriers for recording knowledge (GB 3792+0-83, general rules for literature description). It consists of four parts: document information, document carrier, symbol system and recording mode. From different angles, literature can be divided into different types: from the form of publication or processing, literature can be divided into books, periodicals, patent literature, standard literature, conference literature, product samples, archival literature, "gray" literature and so on. From the level of document processing, it can be divided into primary documents, secondary documents and tertiary documents. With the rapid development of modern science and technology, the quantity and quality of documents are constantly improving, which provides a good foundation for people to use this method to carry out research activities. (1) The required literature should be selected appropriately. Now is an era of information expansion, and a large number of documents can provide us with corresponding information. In order to find documents suitable for the research objectives from a large number of documents, we should pay special attention to the selection of first-hand original materials, relatively authoritative information sources and complete and systematic literature materials. For historical documents, we should pay attention to textual research, discard the false and retain the true, discard the rough and select the fine, and do not stick to the rules. For example, the French scholar George Lefevere (1874- 1956) consulted 1066 notaries' records and a large number of other documents, wrote the book Farmers in Noel County during the French Revolution, compiled more than 200 pages of statistical charts, and profoundly showed the rural situation in Noel County during the Great Revolution.
(2) appropriately determine the observation items of a certain kind of literature. There is a lot of information in the literature, so it is necessary to determine which items to observe and avoid interference from other information. If you want to study the authors of an academic journal, you only need to find the information of the relevant authors from the literature, and you don't need to collect other information. This can improve the efficiency of research work and ensure that the information you want to obtain is not polluted by other information.
(3) The distribution table of observation items shall be listed. In order to have a good idea of observation, we usually need to determine some standards and items to be obtained before investigating the literature to guide future observation activities. For example, taking the relevant literature of A Dream of Red Mansions as the research object, we can take the time when the literature was introduced as the bus, and take the literature of each era as the observation content according to the category, author, publication time, main viewpoints, publication language, author's region, author's work organization, document title and other items. Through the classification, summary and statistics of information, we can find the research status and development trend of A Dream of Red Mansions.
(4) Summarize, count and analyze relevant facts, terms and data. Because people study from different angles and researchers have different understandings, there are often different explanations for some things, thus forming many concepts. For example, for a concept, due to different levels of people's understanding in different times, scholars from various disciplines will put forward various definitions from their own research. Literature research takes many definitions as the basis of research, summarizes and analyzes them, and finds out their similarities.
(5) Relevant opinions and conclusions should be drawn on the basis of analysis. Through the study of a large number of documents, we can get a lot of purposeful data. On the basis of these data, we should use analogy, induction, statistics and other methods to find the content of * * * from common facts and data, and then use logical methods to abstract it to a certain extent, and finally form a certain theoretical point of view or conclusion. "Nanjing University's Contribution to Western Library Science and China Localization" [2] Through the investigation, analysis and summary of a large number of documents, it is found that Nanjing University has made great achievements in offering library science-related courses, establishing the earliest library science specialty, establishing the earliest library science journal, establishing the China Library Association and cultivating library science professionals. It has played an important role in the localization of western library science in China.