Coursework assignment: Literature review

This assignment is weighted at 50% of the final mark for the module.
The purpose of this coursework is to help you to:

• reflect on your research interests in the area of data science and select a topic to focus on
•  identify specific problems within the topic of your interest
• evaluate suitability of the identified problems for a MSc data science project
• get practical experience in analysing relevant sources of information
• get practical experience in writing a literature review report.

For the coursework, please select a topic relevant to data science. You may wish to select one of those we discuss in this module: business analytics, artificial intelligence, text mining, and knowledge discovery in databases. You are not limited to those topics and free to consider a wide range of topics.

Then, narrow down the selected topic to particular issues, gaps or areas of discussions. For example, you can narrow the topic of artificial intelligence to the issues of explainability and interpretability. You can select the topic of text mining and narrow it down to the problem of resolving ambiguity of natural language expressions. Whatever you choose, it should be of interest to you and have a potential to lead to a project you may wish to carry out in the near future.

You will be then more motivated to explore in depth the selected topic. A literature review is at the heart of any good research project. It helps to identify unsolved problems and existing gaps, leads to formulating new tasks, and shows directions for solving those tasks on the basis of achievements of previous research. The literature review will help you to:

• understand the history of the subject that you intend to investigate
• understand the significance of the studies that have been already conducted in the field
• become familiar with the vocabulary and various definitions
• become aware of the current research and debates
• identify gaps in the literature and neglected areas.

You should synthesise the various views and opinions expressed by various scholars. We recommend basing your literature review on a wide range of sources, including books, journal articles, conference proceedings,unpublished articles (e.g. preprints submitted to online archives), news articles, blogs.

The area of data science is evolving fast, and we advise you to focus on recent publications. Your reference list should include 50-70% of references within the last three-five years. Writing a literature review may be time consuming. Therefore, you are welcome to use the material you have prepared during your peer review exercise. You are free to choose not to do so and to focus on an entirely different topic or problem.

The structure of a literature review report

The title of your report

Abstract [15 point]

Give a summary of your literature review. Try to include the following:

• a high-level description of the topic area
• an overview of the problem studied and why this is interesting/relevant/important
• a high-level description of the approach taken
•  a summary of the conclusions.
• If possible, try not to exceed 500 words.

A good abstract is difficult to write and can only be completed after the full report has been written. It represents a brief summary of the review you have carried out and its results. By summarising the results of your review, you allow other people to get an idea of what was accomplished without having to read through the whole report. Otherscholars can read an abstract to decide if looking at the full report will be worthwhile. The abstract should provide sufficient information about the review that reading the full report is not necessary, although your markers will read the full report.