“Don’t reinvent the wheel; just realign it.” — Anthony J. D’Angelo
When starting your research project (especially one connected to a company project), you want to frame the problem accurately and comprehensively. A good literature review does three jobs: it summarizes what is known, clarifies the exact variant you are studying, and motivates why your thesis is needed (your gap and contribution).
- A literature table that compares the most relevant papers (variant, assumptions, method, data, evaluation).
- A short “gap statement” you can reuse in your Introduction (“Existing work does X; however, Y remains underexplored …”).
Tools
Use these tools together to (1) find a few strong “anchor papers”, (2) expand outward via citations and related-work graphs, and (3) keep your reading notes and PDFs organized from day one.
- Research Rabbit: explore citation networks and clusters around your anchor papers (quickly identify the “core” papers and adjacent subtopics).
- Google Scholar: fast way to find “who cited this?”, locate newer follow-up work, and discover alternative names/keywords for the same concept (use Cited by and Related articles).
- Zotero: manage references and PDFs, tag and search your library, and store notes per paper. Research Rabbit can export directly to Zotero (see the Research Rabbit help docs: https://researchrabbitapp.com/help).
The Wall and the Brick
Think of the literature review as the foundation of your research, much like a wall built from academic papers.
In the introduction, you identify a gap in this wall, highlighting where existing literature falls short. Your methodology then introduces the “brick” that fills this gap. While it is common to hint at this solution in the literature review, detailed information about the brick should be reserved for the methodology section. This structure ensures your research both acknowledges the existing academic landscape and clearly delineates your contribution to it.
Literature Review vs. Literature Overview
A common mistake is conflating literature reviews with literature overviews. Here are some key differences:
- The literature review is a critical analysis of existing research, summarizing and synthesizing findings to identify gaps and establish the context for your study.
- Literature overviews are descriptive summaries of existing research, providing a broad understanding of the topic without necessarily identifying gaps or offering critical evaluations.
If you find yourself repeating what has already been stated in other reviews or textbooks, you are likely writing an overview rather than a literature review. Instead, focus on synthesizing existing research and identifying gaps that your study will address.
Research Gaps
Write your review so it naturally ends with a clear gap statement and a defensible contribution. Your gap should be specific to your constraints (data availability, uncertainty, timeline, deployment context), and it should lead to a concrete “brick”: method/design choices and an evaluation plan.
Use the dedicated guide on Research Gaps for templates and examples you can reuse in your Introduction.
Parts of a Literature Review
Below I list common “moves” you can borrow. The excerpts are there to give you phrasing patterns—not to be copied verbatim.
Scoping the Literature Review
Scope your literature review by focusing on specific aspects of the problem you are addressing. This helps narrow down the vast amount of literature to what is most relevant to your research. For example, you might focus on specific methodologies, algorithms, or applications within your field:
- “The literature survey below will pay special attention to how the above-mentioned two central issues, the solution of the pricing problem in BCP algorithms and the solution of the interdependence problem in (local search) heuristics, are addressed. Other potentially promising ways for solving VRPMSs and their potential advantages and drawbacks will nevertheless be outlined.”
- “This section provides an overview of various VRP modelling methodologies supported by data analytics and machine learning. In particular, we focus on modelling techniques for handling uncertain, incomplete, imprecise, or ambiguous data in VRPs, including stochastic programming, robust optimisation, chance-constrained programming, and data forecasting.”
Referring to Other Literature Reviews
You can refer to other literature reviews to provide context or background for your work. You do not need to repeat what has already been said. Instead, refer to existing reviews and focus on specific aspects relevant to your research. For example:
- “The readers may refer to Elshaer and Awad (2020) for a comprehensive review of metaheuristics for VRP. This review focuses on machine learning-assisted VRP algorithms, and Section 4 discusses all relevant papers.”
- “For a comprehensive review of robust optimisation, we refer interested readers to Bertsimas, Brown, and Caramanis (2010).”
- “A good overview of such algorithms is provided by Molenbruch et al. (2017), who note that ‘in the standard problem, operational costs are minimized, subject to full demand satisfaction and service level requirements,’ closely relating to our study presented below.”
- “Details on methods used to operationalize this controller have been extensively documented in several publications (Oh et al., 2020a; Nahmias-Biran et al., 2019; Basu et al., 2018), to which we refer interested readers.”
- “The synthetic population for 2030 (6.7 million individuals) was generated using a Bayesian approach (Sun and Erath, 2015) based on socio-economic data, land-use data, and relevant control totals (see also Zhu and Ferreira, 2014, for more details on the population synthesis).”
- “The calibration and validation also included matching simulated outputs to observed screen-line counts, public transit smart card data, and network travel times (for more details on the model calibration, readers are referred to Oh et al., 2020a).”
- The presentation of the material assumes familiarity with vehicle routing problems (capacitated VRP, VRP with time windows VRPTW, pickup-and-delivery problem with time windows, capacitated arc routing problem, dial-a-ride problem, etc.) and with the standard modeling and exact and heuristic solution methodologies (mixed-integer programming, branch/cut/price, local/neighborhood search, metaheuristics). If this is not the case, the reader is referred to Toth and Vigo (2002); Golden, Raghavan, and Wasil (2008); Desaulniers, Desrosiers, and Solomon (2005); Funke, Grünert, and Irnich (2005); Røpke (2005); and Gendreau and Potvin (2010).
- For more information on the split delivery VRP or pickup-and-delivery problem (PDP), the reader is referred to Hooker and Natraj (1995); Chen, Golden, and Wasil (2007); Archetti and Speranza (2008); Nowak, Ergun, and White (2008); Schönberger et al. (2009); Desaulniers (2010); Derigs, Li, and Vogel (2010); and Hennig (2010).
Literature Review Table
To effectively summarize the literature, consider creating a table categorizing existing research. This table should include key characteristics of each variant, relevant papers, and methods used. Readers appreciate this structured approach, as it allows quick comprehension of existing research.
To ensure your table is useful, search for similar tables in existing literature relevant to your topic. Identify key characteristics typically relevant to your field, which can then become columns or sections in your table.
Regardless of the specific characteristics chosen:
- Ensure the table is clear and concise, allowing easy comparison of studies. If the table becomes too large, split it into multiple tables (e.g., one per variant category).
- Explain the table in the text before presenting it, clarifying its purpose and interpretation. Each column and column value should be clearly defined and explained. Refer to previous explanations when appropriate.
- Clearly state the reasons for including each paper in the table. Readers should understand why specific papers were chosen and how they contribute to overall topic understanding.
- Make the table self-contained and understandable without referencing back to the text. Include all necessary information within the table itself. Abbreviations used in the table should be defined:
- In the table header (e.g., your header could be “(D)ynamic/(S)tatic” with values as “D” or “S”).
- In a footnote below the table.
- In the caption of the table.
- Add a final row to the table titled “This Thesis”/“Current Study”/“This Paper” to clearly indicate how your work fits into the existing literature.
- The first column should be the author(s) and year of publication (
\citetin LaTeX). - Order papers to easily identify key characteristics quickly. Papers discussing the same variant as your thesis and those closer to your methodology should be grouped together near your thesis at the bottom of the table.
- Cite all papers included in the table within the text beforehand. This ensures readers are familiar with the papers and can refer to them for detailed information.
Tip: A good example of a concise literature review table can be found in Ulmer, Goodson, Mattfeld, & Thomas (2019).
Concluding the Review
End your review with a concise summary clearly stating your research variant, emphasizing your contribution, and positioning it within existing literature:
“In this study, we specifically focus on the Dynamic Vehicle Routing Problem with Time Windows (DVRPTW). Our research addresses the integration of real-time stochastic demand forecasting with robust optimisation techniques.”
Examples of Literature Reviews
- Shared autonomous vehicle services: a comprehensive review (Narayanan, Chaniotakis, & Antoniou, 2020).
- A survey of dial-a-ride problems: literature review and recent developments (Ho et al., 2018).
- Optimization problems for machine learning: a survey (Gambella, Ghaddar, & Naoum-Sawaya, 2021).