How To Organize A Research Paper

(Adapted from "How to Organize your Thesis" by Prof. John W. Chinneck)

What Graduate Research is All About

The distinguishing mark of graduate research is an original contribution to knowledge. A research paper is a formal document whose sole purpose is to prove that you have made an original contribution to knowledge. To this end, your paper must show two important things:

  1. you have identified a worthwhile problem or question which has not been previously answered,

  2. you have solved the problem or answered the question.

A Generic Research Paper Skeleton

(1) Title & Author(s)

(2) Abstract

A very concise summary of the problem addressed and solution presented in this paper.

(3) Introduction

This is a general introduction to what the paper is all about -- it is not just a description of the contents of each section. Provide an introduction to the research area and summarize your research question (you will be stating the question in detail later). Provide some of the reasons why it is a worthwhile question. Give an overview of your main results. This is a birds-eye view of the answers to the main questions addressed in the paper.

(4) Literature Review

Here you review the state-of-the-art relevant to your paper. The idea is to present the major ideas in the state of the art right up to, but not including, your own personal brilliant ideas. Do NOT just list the papers you found and explain what each paper does. Some papers may be improvements on previous papers. So only the better paper should be explained and the previous paper only needs to be listed. Organize this section by idea, and not by author or by publication.

For example if your paper is about Parallel Sorting Algorithms, you might organize your Literature Review subsections around different approaches or cases:

  • Parallel sorting for MIMD, shared memory architectures

  • Parallel sorting for MIMD, distributed memory architectures

  • Parallel sorting for SIMD, shared memory architectures

The main goal here is to explain the state-of-the-art in the field.

(5) Research Question or Problem Statement

Computer Scientists and Engineers tend to refer to a "problem" to be solved where other disciplines talk in terms of a "question" to be answered. In either case, this section has three main parts:

  • a concise statement of the question that your paper tackles

  • justification, by direct reference to your Literature Review, that your question is previously unanswered

  • discussion of why it is worthwhile to answer this question.

(5) Description Of How You Solved the Problem or Answered the Question

This part of the paper is much more free-form. It may have one or several sections and subsections. But it all has only one purpose: to convince the reader that you answered the question or solved the problem that you set for yourself. In this section you will for example present new algorithms you developed and your implementation of these new algorithms.

(6) Experimental Evaluation

This section is not mandatory for all papers (for example theory papers) but typically required for papers in the field of parallel computing. After all, parallel computing is all about compute performance. Here you present performance data obtained from e.g. running your newly developed algorithms and code on a parallel machine using some benchmark input data. Typically, you need to describe the parallel machine you used and the data that you used as input. The main results are usually performance graphs, typically speedup curves. You want to evaluate your code on different input data sets highlighting the strengths and weaknesses of your code. Don't just use best case scenarios. People will call you on that. Discuss the results obtained.

(7) Conclusions

You generally cover three things in the Conclusions section.

  1. Conclusions

  2. Summary of Contributions

  3. Future Research

Conclusions are not a rambling summary of the paper: they are short, concise statements of the inferences that you have made because of your work. All conclusions should be directly related to the research question.

The Summary of Contributions will be much sought and carefully read by the readers. Here you list the contributions of new knowledge that your paper makes. Of course, the paper itself must substantiate any claims made here. There is often some overlap with the Conclusions, but that's okay.

The Future Research should indicate interesting new problems arising from your work. No paper ever solves everything. In fact, the best research papers lead to new research questions for other researchers to work on.

(8) References

The list of references is closely tied to the review of the Literature Review. Most readers scan your list of references looking for the important works in the field, so make sure they are listed in your Literature Review. All references given must be referred to in the main body of the paper. Organize the list of references alphabetically by author surname.