LITERATURE REVIEW TEMPLATE
NOTE: Please use the “Grammar Check” link in the left side bar in Interact to check the grammar of your final report before submitting it.
Important information to read before you start your Literature report:
A literature review describes and evaluates the research you have read about the specific topic that is of interest to you (very detailed). It also includes a critical analysis of the relationship among different works. All this, you need to also relate to your work.
Your Literature Review Assignment should be less than 2500 words and have the following sections:
Abstract: (altogether ~200 words)
Give a brief overview of the aims, methods and expected findings of your work, and your contribution in this project. Write each of these points as small paragraph.
1. Introduction: (altogether 300~400 words)
* Give a definition of the topic you are researching (a description, an argument or a proposal)
• Tell your reader what is of interest to you for your research and state the problem (the narrow field you have researched, i.e. automated feedback for tertiary students).)
• Tell your reader what it is NOT of interest to you for your research (i.e. you are NOT doing research on primary or secondary students)
• State the overall findings from your review (i.e. what do most sources conclude).
2. Literature Review: (1600 – 1800 words)
• Describe the field of research and show your reader how you will organise your research (in sections with headings)
• Write your 12 reviews, using the following guide:
1. Organise the 12 reviews in blocks (depending on what is important for you) – you may have noticed categories some research papers fall into.
? By technology
? By results
? etc.
? (Pl do not discuss paper by paper)
2. Check all papers to see what you can describe together.
3. Describe the research (what they did; how they did it; what technology they used) + evaluate the contribution the work makes to the field (i.e. why is this work important; how does it differ from other work) + show the limitations of the research (what did the researchers identify as limitations)
+ show what you think are omissions, gaps or limitations (you know this topic now; you should able to identify if they have not done something: gaps and limitations are the justification for doing this research) + show how the research is valuable for your project (link it to your work). See 2nd page for the sample
4. Identify relationships among studies in the same group/block:
Note relationships among studies (i.e. this work achieved higher accuracy than other studies using the ROTTO algorithm because …). Identify important studies (landmark studies) that led to subsequent studies in the same area.
5. Keep your review focused on your topic: make sure that the articles you have found are relevant and directly related to your topic.
6. If you include tables, diagrams, flowcharts or formulas as part of your review, each of them should be given a figure/ table number and that number must be cited in the main text with a detailed explanation (below the figure).
• Add and explain the Component and Classification tables. Need to explain why you selected those columns and the importance of these tables.
Sample review group (see below for codes A – H)
Group/Heading
Research using convolution Neural Networks (CNN)
[A] Convolutional neural networks (CNN) have been used to achieve a more fine-grained analysis in a range of research works. They are generally effective for smaller corpora (Hadad et al., 2016; Colton et al., 2015). However, processing a large corpus is still computationally expensive due to the high processing time. [B] To overcome this, Abasir et al. (2017) used CNN to train a Natural Language Processing (NLP) system creating a RoTT1 algorithm to reduce the training time of a large corpus by several seconds. [C] Lowering training time is important since it lowers computational cost [D] and even though the improvement was not of magnitude, no other group of researchers has been able to achieve this. It points to the fact that the RoTT1 algorithm has potential. [E] In addition to the size of the decrease, the researchers identified as limitation that the overall impact of this reduction is minimal as the data preparation time increased significantly, offsetting the time gains from training. [F] Despite these issues which require further research, the RoTT1 algorithm developed by the researchers is still of interest to the proposed work as it has proven to reduce processing time which is one of the aims of this research. [G] The RoTT1 algorithm has since also been used by other researchers. An algorithm based on similar principles and CNN were also the basis of an object detection system by Compten and Wells (2018). However, although this adaptation of RoTT1 is interesting, it is field-specific, and the concept may not be suitable for adaptation because the development of a new algorithm is not the focus of this work. It is, therefore, of no further interest. [H] However, of considerable value for our work is the research by Blander et al (2019) which focused on [A] compression of corpora by … Blander at a l (2019) used a multi-layer CNN with the aim adding layers that would sort and condense the corpus (next review in this same category, repeating all sections A, B, C, D, E, F)
Article adapted for educational purposes from Ho et al. (2018) with permission of the authors.
[A] State the special area this paper is based on [get additional references from general research or from the paper you have reviewed but not used].
[B] Introduce the work you want to discuss and what they did.
[C] Tell your reader why their work is important.
[D] Draw a comparison about this work with existing research.
[E] The limitations the researchers identified themselves.
[F] State why this work is of value to your own research
[G] How has research been used in the field?
[H] Create a link to the next valuable piece of research (i.e. one of the 12) in the same category.
• Frequently link the discussed research findings to the research question stated in the introduction. These links create a thread of coherence throughout your review article.
3. Conclusion (150-200 words)
Summarise the key findings from the literature review (positive); restate the limitations; based on these limitations, justify your research proposal (i.e. discuss the gap/niche you identified).
4. References (~ 75 articles)
Please note that you need to add all 30-35 references you have collected in your weekly submissions and the other additional 2 references you have given for each journal paper. That means you should have a reference list of 75. Please do the citation of all these papers in the main document (mindful of the comments on referencing above).
PLEASE note that you do not have to propose your OWN systems; just provide a diagram of the topic or any landmark solutions.