How to write a good research paper

Being a PhD, I have published quite a few scientific papers in different conferences. Here, I will give hints on how to write a good paper. So, let's get started.

image source

Organization of the paper

All the scientific papers are organized in almost exactly the same way. You start with an abstract, then have the introduction, followed by background/related work, then your design and implementation, followed by results and finally the conclusion.

Next, I'll write about how you should write each of these parts.

---

Abstract

Abstract is a kind of a super short summary of your whole paper. Here you describe what your basic idea is, and how it is better than the existing ideas (which are also called state-of-the-art). You then back up your method by showing how much improvement you got as compared to the state-of-the-art methods.

Introduction

In the introduction, you basically introduce the reader to your ideas in a succinct manner. You also state out explicitly what are the contributions of your paper, and how it improves upon existing solutions. Lastly, you tell the reader about how your paper is organized, and list out the parts of your paper.

Background/related work

Before you describe the design and implementation of your own method, you must give a background on what problem you are trying to solve, and what has been done so far in this field by other researchers. Also briefly talk about how your implementation improves upon the state-of-the-art approaches.

Design and implementation

In this part, you describe in detail the design and implementation of your method. Rather than going into the details immediately, you must first give a high level introduction of your design, and from thereon go further into details. In this way, it would be much easier for the reader to grasp your idea. Also put figures generously to make the reader help understand better.

Results

Here you show the empirical results that you got from your method. Not only you should show your own results, but also compare your results with the start of the art, because without any comparison, it would be impossible to know whether your method is any better than the state of the art. Besides listing your results in tables, it is also a good idea to show them in graph forms.

Conclusion

Here again you give a brief summary of your work, and state the improvement you got over the state of the art methods.

---

Besides these parts, your paper could also include Acknowledgements. And of course you would also have references at the end, for the papers that you cited.

Note that this kind of organization is not only helpful for writing scientific papers, but also writing in general.