How to read documents

One

1. From point to area. Select the questions and hot spots in practice, and retrieve a wide range of documents from a small branch. Generally, there are more than 20 articles in the near future. The reason why I don't need to care about it three years ago is that the knowledge is updated very quickly, and most of the information I can find on the Internet is the full text of the past few years. Learn how others discover and solve problems. I understand the common views and differences on this issue. Then, expand and, based on the interests and purpose of the research, find out the fields of research: Who has been cited for many times, who has the most up-to-date articles and the most enlightening ones. Go to the library to find his article. Gradually expand your horizons and build your own professional knowledge structure and views.
2. From miscellaneous to excellent. With a certain knowledge base, you must have personal judgment on complicated documents. Tracking the research progress of a topic and an expert, comparing the development of arguments on the same topic, mastering its new methods or conclusions, or paying attention to the changes in the author's views, and exploring the reasons. Cultivate personal academic accomplishments. For high-quality and high-level journals, you can browse regularly to learn about academic progress and hot topics. Based on your personal interests and work progress, read new works one by one.
3. Poor memory. Whether it's a bit of discovery or a spark of thinking at work, you should write it down. When I was in contact with Academician WANG zhongcheng and academician Gu Yudong, I found that they all had the habit of memorizing cards. Case Follow-up, literature views, and so on. When writing articles, they are all ready-made materials. Now we have computers, but writing literature reviews is a good way to improve the knowledge structure. Keep down your arguments and personal experiences at any time, and you will get twice the result with half the effort. Whether written on paper or recorded on a computer, there should be a notebook and it should be organized frequently.
4. For the downloaded documents, it is necessary to establish a special category by time based on their content. Which need to be carefully read and saved, which are of little use, to be deleted, and which need to be read but not yet read. You can find them in time later.
5. Learn every day. Documents are available every day. If you only serve as a collector, you will lose the significance of the study. Download aims to learn. Learn the methods and knowledge in the professional field through reading. As long as you keep learning, you will build your own knowledge architecture. It is easy to use.

II.

For beginners who are new to a field, they must read a large number of documents to grasp the dynamics and directions of the field.
I remember a foreign graduate student saying that at first the tutor asked him to read a large number of documents and set the number of documents every day, as if it were 100 articles? As I have just been in touch with this field, I have no idea about many problems, so it is very difficult to read and understand many contents. Ask the tutor, but be told that you only need to read enough data every day. Later, with the increase in the reading volume, I finally integrated and understood the mentor's method.
Therefore, I think for new users, we should pay attention to the number of reading documents, and the accumulation of documents increases from quantitative change to qualitative change.
In addition, each author's research methods are somewhat different. After reading more, he will gradually compare the advantages and disadvantages of the research methods, which will be of great benefit to his future research.
In fact, due to the rapid progress of science and technology, many new technologies and new ideas are constantly emerging in even the fields that you are engaged in. Therefore, if you are a veteran, if you are too lazy to update your knowledge, it will soon lag behind.

3

Today, with so many articles, I feel like an increasing number of posts.
1. Grasp the dynamics. after entering the park every day, I will first browse new posts since my last login. If you are interested, I will check them again. If necessary, I will save them or add them to my favorites. This is roughly equivalent to the magazine directory, this ensures that nothing new will be missed.
2. Let's take a look at the best posts, popular posts, and encrypted posts of interest in the forum. This is equivalent to the reviews, comments, or cutting edge (JI) of brick magazine.
3. There are some good things not only in sections of interest, but also in other sections. When you are blank, check the excellent posts, popular posts, and encrypted posts of other sections, this is equivalent to a review, comment, or cutting edge (JI) of brick magazine.
4. Use the search function to search for posts with specific keywords.
5. As you become familiar with the field, especially those who are familiar with masters, you can sometimes look at the posts of these authors.
6. In today's information, no one dares to say that they are already super-class and do not need to accept new things. No one can be familiar with all fields.

4th

Academician Zou chenglu wrote about how he read the literature:
No matter where the question comes from, it is necessary to closely track the current trends in the field of science. Starting from the Postgraduate age, my tutor taught me how to use my classmates as an example. I developed the habit of going to the library to browse the latest journals every week. If something happens, you must make it up next week. At that time, I had a small Record Book, logging on to all the important publications of this major. After reading each record, I will make a record and never miss the first record until today. Now you can read the directories and summaries of all important publications online, which makes it easier. Master the literature, make a comprehensive analysis of the literature, and evaluate the literature from a critical perspective, it also extracts useful and correct information from it to guide future research. This is an essential capability for independent scientists.
When reading the literature to track the current development trend, you must make full use of your judgment and never blindly follow. Even well-known scientists and textbooks sometimes make mistakes. The ancients said well: "It is better to believe in books than to have no books ". When tracking the important direction of current development, remember that the problems you see will also be seen by others. The more important the problem is, the more intense the competition. When the research conditions are not as good as people, without innovative research ideas, it is impossible for a unique research solution to succeed beyond others. Although some important topics have emerged internationally, it seems very serious in China. Lack of innovative ideas, and one-sided pursuit of hot spots, is an inevitable line of failure, at most can only be to icing on the cake for others' achievements, or do some minor repairs and supplements. The key lies in your own innovative ideas. Where is the idea of innovation? Although an important idea of innovation has been generated in the history of science, it is not uncommon, but far more common is that genius is hard-working, from accumulation, innovation can be personal accumulation or long-term accumulation of your organization. This is the Wang stove principle mentioned above, which is also the reason why the Nobel Prize often exists in a few units. Only by hard work can we continue to accumulate excellent work, and then we can gradually produce innovative ideas in our work. Only when others cannot plagiarize innovative ideas can we make breakthroughs on major issues. In a fire-burning fire, there will be continuous accumulation of excellent work, the emergence of excellent talents, and the continuous interaction of innovative ideas and talents, mutual inspiration and mutual encouragement, it will continue to create new breakthroughs.

5th

Professor at the mit ai lab and students have summarized how to read the paper!
Reading a thesis is a skill that requires practice. It is impossible to fully read all the papers.
The thesis can be divided into three stages:
The first stage is to check whether there are any interesting things in the paper. The AI thesis contains summaries, which may contain content, but may not be well summarized or summarized. Therefore, you need to skip reading, understand what the author has done. The table of contents, conclusion, and introduction are three key points. If none of these methods work, you have to browse them in sequence. Once you have figured out the rough idea and innovation points of the paper, you can decide whether to proceed to the second stage.
In the second stage, we need to find out what the paper really has. Many 15-page papers can be rewritten to about one page. Therefore, you need to look for exciting places, which are often hidden somewhere. The areas of interest discovered by the author in his work may not be yours, and vice versa.
Finally, if you think this paper is indeed valuable, you can go back to the topic for intensive reading. When reading a thesis, remember the question: "How should I use this paper ?" "Is it really like what the author claims ?" "If ...... What will happen ?". Understanding the conclusions of this paper is not equivalent to understanding the paper. To understand the thesis, we need to understand the purpose of the paper, the author's choice (many of which are implicit), and whether assumptions and formalization are feasible. What direction has the paper pointed out, what are the problems in the fields involved in this paper, what are the patterns of difficulties that the author has been studying, what are the strategic points of view expressed in this paper, and so on.

6th

1. Read the abstract in most articles and the full text in a few articles
I have mastered some full-text query skills, and I often enjoy the full text, so that I don't have time to read the content of the article, so I prefer to read the abstract. There are not many useful full texts. It is a waste to pursue the full text. Of course, reading the abstract is not correct.
2. concentrated time to view documents
Always forget. The more time the document is viewed, the more time the document is wasted. Concentrated time is more easily linked to form an overall impression.
3. Record and Mark
Documents copied or printed are directly marked or annotated with a pen. Document in PDF or HTML format can be highlighted or changed in the text color in the editor. This is another important way to avoid wasting time. Otherwise, you will not see it.
4. Read the articles to be referenced.
The conversion causes are too many examples.
5. Pay attention to the reference value of the article.
The influence factors of publications and the number of cited articles can reflect the reference value of the articles. However, pay attention to how other articles cited in this article evaluate this article: Support, opposition, supplement, and correction.