Preface
After studying for graduate school, I have been thinking about a question: How should I choose to read scientifically in the face of massive literature? How to quickly filter out the documents you want?
After searching thousands of documents through one or a few keywords, read one by one? Not too possible. Time-consuming and labor-intensive. After reading a few highly cited papers, I still don’t know what I’m reading. See the roundup? It’s a good idea, but the review may not be available, and it may not be newer.
I found that the efficiency of reading a document is too low. Facing a newer thing, a document is like a brick. I read a document and it is equivalent to obtaining a brick. I want to build a brick with bricks. Building, how many documents do you have to read? When can we build the entire knowledge framework?
So I was wondering if there is anything that can help me quickly build a knowledge framework for a research field or direction? After building a frame, I read a document to know which floor or room it is located on, and I can put it directly into my knowledge frame.
So what is the knowledge framework for the research field? The following questions may help to understand. such as:
- Who is studying in which institution in which country?
- What are the subdivisions of this research field?
- What is the development history of this research field? How did the research direction evolve?
- Which papers are landmark papers worth reading?
- Which papers influence the follow-up research direction?
- What are the most popular research directions now?
- Where or what is the future research direction?
- Which journals are these papers published in?
Etc., etc
As a graduate student, if you don't know the evolution history of the discipline or research field or research direction, you don't know who your colleagues are, and you don't know the research trend, you can say it is unqualified.
Scientific knowledge graph analysis-a new solution
In fact, to be honest, it is not a new solution. The scientific knowledge graph is a concept in the field of library and information science. It has been applied in the last century, but I just discovered that this thing can be used for document analysis.
basic concepts
The Mapping Knowledge Domain is a graph showing the relationship between the development process and structure of scientific knowledge. Express the laws of scientific development in mathematical equations and present them in the form of maps. Taking scientific knowledge as the object of metrological research belongs to the category of scientometrics.
In China, the scientific knowledge graph is often referred to as the knowledge graph, but its connotation is different from the knowledge graph proposed by Google. The scientific knowledge graph is a concept in the field of library and information science, which focuses on the visualization of knowledge units. Connection, Google Knowledge Graph is a concept in the computer field, focusing on describing concepts, entities and their relationships in the objective world in a structured data form (such as RDF).
effect
- Reveal the history of scientific knowledge development
- Discover the research hotspots, frontiers and trends of subject content
- Reveal the relationship between disciplines
- Researchers, institutions, cooperation between countries
analyzing tool
- CiteSpace-comprehensive analysis
- HisCite-Citation History Visualization
- Bibexcel-field extraction / generation of co-occurrence matrix
- VOSviewer-keyword co-occurrence analysis
- BICOMB-co-occurrence analysis
- SATI——Analysis of Bibliographic Information
- SPSS-clustering/multidimensional scale analysis
- Ucinet/Pajek-social network analysis visualization
- TDA——Patent Map Analysis
- Areak——Patent Map Analysis
- NetDraw-social network analysis visualization
In fact, there are still a lot of software, which will not be detailed here.
What needs special reminder is that a tool is a tool, only an aid, not a substitute for literature reading. Literature reading is still the book. Don't do things that put the cart before the horse.
------2019.10.30 update-----------
citespace
Next, I will learn citespace, and I will summarize some of the experience, problems and solutions during the learning process here for your reference, and also for my own backup and follow-up search.
Download and install
download
Citespace relies on the Java environment, so you need to download Java before installation.
Java download address: https://www.java.com/zh_CN/download/
Java can be installed directly after downloading, there is nothing to pay special attention to, it is very simple.
Citespace official website: http://cluster.cis.drexel.edu/~cchen/citespace/
Find download in the official website and click, then jump to a new page, and then click download, it will jump to the sourceforge website to download.
Or directly click citespace download to jump.
After downloading, you can install it.
Citespace installation
Note: Software installation is time-consuming and variable, which has a lot to do with network speed. My internet speed is poor and it took about 35 minutes.
The installation process is divided into 2 steps:
- Initialization of citespace, this step is mainly to create a citespace folder
- Download the WOS Demo.
1. Run in administrator mode by right-clicking and initialize the software. The
following interface will appear.
Wait for the progress bar to reach 100%, and then the new folder for initialization will pop out, as shown below.
2. Click "OK" to download WOS Demo, the process is as shown below:
When the progress bar reaches 100%, the downloaded WOS Demo file will pop up, as shown in the figure below:
Click "OK" and wait a few seconds to enter the citespace interface.
Only after completing the above installation steps can citespace be fully installed. However, the software can be used normally without downloading the Demo.
HowNet Data Download
The download of HowNet literature data is very simple. I will not go into details here, but only record a few points for attention.
- Pay attention to the differences in search fields when searching. Use keywords, topics, and title fields for the same search content to display different results.
- Select Refworks format when downloading data.
- The name of the downloaded text file must start with "download".