How to Write a Good Survey paper
- 1. How to Write a Good Survey Paper A survey paper classify existing different approaches to the problem, using a well set of classification criteria. The main idea behind the survey papers is to build a framework of a research topic based on existing literature. Survey paper does not publish new research. Surveying research papers generally mean to collect data and results from other research papers. As per the ACM Computing Surveys Editorial Charter “A survey article assumes a general knowledge of the area; it emphasizes the classification of the existing literature, developing a perspective on the area, and evaluating trends”1 . A good survey paper exposes research topic that needs more research focus. A survey paper mainly compile the information that is currently known about the topic. It provides basic knowledge to new researchers. It also qualitatively and quantitatively validates information to researchers. Selecting a topic for survey is a challenging task. To make a good survey topic, author should satisfy the following requirements: 1. It is an emerging field and popularity of the field will grow (over time). 2. Papers with new algorithms/approaches does exist. 3. A survey paper does not exist on the survey topic. Author can start with reading general topics before moving to pick-up relevant topic. It is advisable that author should read 20 to 40 (at least) well written papers before start writing a good survey paper. Citation count and locality principle for papers could be a good idea to select good papers. Before writing, author must develop good understanding of the field. Author must completely be aware of the main research themes, trends, challenges/issues, and results of that field. The author must add his/her research experiences related to this research topic (if possible). The author must produce the essence of the topic in a legitimate way. To differentiate between approaches, author must present each approach using same template and same type of figures (for easier understanding). It is recommended to have examples, classification, figures, summery or tables to provide readers a good overview of the topic. Author should explain example and its potentials using his/her understanding of the field. An example can be explained in detailed by using seven Ws (Who, What, When, Where, Why, for Whom, hoW). To classify available approaches/methodologies author must decide what should be the main criteria. It is always recommended to find the reason that why this classification criteria is a contribution to the knowledge domain. It is desirable to present classifications with examples. Figures must be clear and self-explanatory. e.g, figures can be used to explain examples in pictorial format or to present an algorithm in pseudo-code format. Research methodology/strategy must be included to analyze hybrid approaches. It is also desirable that author should discuss plausible solutions of the existing challenges. Information flow in the paper must be well connected which will help to flourish the essence of the paper. In other way, it will increase the visibility of the paper. Before submitting to the journal, it must be reviewed by other peers. Lastly, a good survey paper must adhere the below structure: 1. Introduction to the topic
- 2. 2. Program statement (how to classify and how to survey) 3. Classification 4. Characteristics of classes 5. Presentation of classes 6. Pros and cons 7. Predicted/Emerging trends 8. Conclusion 9. Acknowledgement 10. References Written By: Somnath Mazumdar Dept. of Information Engineering and Mathematics, University of Siena, Italy Reference: 1. http://csur.acm.org/charter.html 2. http://home.etf.rs/~vm/ This work is licensed under a Creative Commons Attribution 4.0 International License.