How to choose a research topic?
I am assuming that the project guide and the area or domain of specialization (e.g. Data security / काया चिकित्सा / Machine leaning) is already decided. “Now what ?”
For the word “research”, we always think about an apple falling on the head or the Eureka moment. Please understand that very few are capable of doing it at the bachelor level, and so I always recommend the students concentrating more on the “applications” related problem statements rather than the “theory or equations”. Now, there are two basic categories of applications-related research problems:
- Manufacturing and service sectors have research needs on continuous basis. The expectations of standards of living create the category of demand-driven research.
- The nature provides abundant opportunities while understanding it. These topics fall into the category of curiosity-driven research.
Talk to your friends, family members, distant relatives – even from different sectors — for understanding what is happening. E.g. sometimes the noise reduction algorithms in automobiles Tata-Nano-car’s engine are also useful in Nadi Tarangini‘s signal processing step. Keep your eyes and brains open with “whys and hows”. Few steps I recommend students to find a research topic are as follows.
[Spend approximately 2 days] First of all, talk to your guide, seniors, library if there is any un-finished work in the domain from the previous batches (typically last 5 years). Is there any future work mentioned in any of those previously submitted documents. If it is there, it’s the best and fastest way to find a topic.
[Spend approximately 4 hours] Select journals (J) and international conferences (C) in your area. Different categories of J&C are:
- Good J&Cs with high impact factor where leaders of the area publish work: These help to understand the current ongoing research across the world.
- Popular J&Cs which have bigger audience: These are usually commercial and contain fast solutions to current topics
- Find out if there are any “TED talks” in the area, usually videos impact more than texts.
- Find out if there are any “Review articles” to understand all aspects of your area in short span of time.
Once you select the J&Cs
3(a) [Spend approximately 6 hours] First read only the titles of 500 articles from last 5 years. Please do not read the articles yet. Once you read the titles, you will start realizing that you like L1 kind of titles.
3(b) [Spend approximately 2 days] Start reading abstracts (free on web) of only L1 articles. You will further realize that you like reading L2 type of abstract more.
3(c) [Spend approximately 2 days] Now go through introductions, findings and discussions of few of the articles from L2.
Please note that this step is not “literature survey”. This step is just to understand the overall research area of your domain and find your likes / dislikes.
4. When you read the above, try to tabulate
— What are the keywords
— What is the novelty
— What are the methods / technologies used
— What is the application
— What is the future scope
In the table, highlight one point in every row. The problem statement for your research will be clear in your mind. Now split that “problem to be solved” into Who, What, Where, When, Why and How question-answers. Usually, the research topic title and scope will be confined to any 2 WH’s of the above 6.
Make sure your research “solves” basic needs in the society. Discuss with guide, seniors and family to finalize the research scope and title.
Next step in the “Simple Steps of (College) Research” series will cover the topic of Literature Survey. Note that the objectives of the research will come after the literature survey.
For regular updates, like and follow: