Ricardo M. Czekster
Home
→  Writing tips
→  PT-FAQ
About
Research
→  FAQ
→  Highlights
FYP
→  Info
→  Projects
→  FYP-FAQ
→  Kit
→  Timeline
MSc.
→  Projects
Programming
→  Useful
→  Snippets
Interests
Projects


Reading list
Rants
Contact

Research FAQ

Questions

• Where to begin?
• What are S.M.A.R.T. goals?
• How to do Science the proper way?
• What are Research Methods?
• How to write my document?
• I don't have any datasets to work, what can I do?
• Do I need to fill out the Ethics forms?
• Can I use LLM/AI-generators (eg, ChatGPT)?
• How to prepare for meetings?
• Which AI/ML algorithm should I employ in my research?


Where to begin?

FIRST of all, to work with research and answer "where to begin?", start by finding a research topic or theme that motivates and interests you.

THEN, you should start by having a 'guiding question', or what we call, a 'research question'. What specifically interests you in this field?

What question are you trying to answer here with supporting scientific literature (based on Google Scholar↗, Scopus, Science Direct↗, Portal ACM - Digital Library↗, or IEEExplore↗, and other scientific venues)?

You should think really hard about this and then let's reconvene.

It was very good that you find the topic, the theme interests you. Now, WITHIN the theme, we need to further clarify what we are trying to discover/unveil/understand, and then how to substantiate with evidence (related work).

Check out this answer at my FYP FAQ.

Read materials about Scientific Research principles.

Go back to questions


What are S.M.A.R.T. goals?

This is used when you are setting your objectives. More information here↗.

They are meant to be:

  • Specific.
    • What needs to be accomplished?
    • Who’s responsible for it?
    • What steps need to be taken to achieve it?
  • Measurable.
    • Quantify your goal.
  • Achievable.
    • Is your objective something you can reasonably accomplish?
  • Relevant.
    • Why are you setting the goal that you're setting?
  • Time-bound.
    • What's your time horizon?
    • When will you start creating and implementing the tasks you've identified?
    • When will you finish?

Go back to questions


How to do Science the proper way?

In a nutshell, you will follow the Scientific Method↗.

 

The term "scientific method" emerged in the 19th century, as a result of significant institutional development of science, and terminologies establishing clear boundaries between science and non-science, such as "scientist" and "pseudoscience", appearing.


Image credit: created by Amy Cowen for Science Buddies / Science Buddies

The idea is to work on defining a problem, looking at what other researchers have done, devising your hypothesis, working on experiments that validate or refuse your hypothesis, keeping on a loop case answers do not satisfy you, until you are satisfied where you communicate your findings to the scientific community or other audiences.


Ideally, your quest will discover things, in other cases, your experimentation might help other researchers from taking the steps you took, and devising new ways of reaching out the objective truth.

Go back to questions


What are Research Methods?

Here's a list of some Research Methods you could use:


[click to enlarge]

Go back to questions


How to write my document?

You will have the following basic structure:


[click to enlarge]

This is not written in stone, you can change here and there. For instance, instead of methods, you could call "Methodology", and also put "Discussions" in a sub-section within "Results", and so on.

Note that in Computer Science, it is common to have a section called "State-of-the-Art" or "Related work" (not in figure), where you put previous related research with comments on limitations/strengths.

Get inspiration from other researcher's work.

Tips:

  • Argue your point, showing limitations and strengths;
  • Discuss in depth, not superficially;
  • Refrain from using/citing blogs, commentary on the Internet, links that won't sustain the test of time;
Go back to questions


I don't have any datasets to work, what can I do?

First things first: you need to get familiar with licensing models (and if they involve human related data, application to Ethics - see related question about this topic).

  • Licenses (more information at this link at Imperial↗):
    • Creative Commons (CC)↗: There are six main license types. The most permissive license - and therefore the license most suited for enabling data sharing and reuse - is the Creative Commons Attribution (CC BY Attribution).
    • MIT↗: permits any person to use, copy, modify, merge, publish distribute, sublicense, and/or sell copies of the software as long as a copy of the license notification is included with any reuse.
    • GNU General Public License↗: users can copy, distribute, and modify the software as long as any modifications are also licensed under the GPL.
    • Apache↗: allows users to use the software for any purpose, to distribute it, to modify it, and to distribute modified versions of the software as long as a copy of the license is redistributed with any modified software.
    • Database Contents License (DbCL) v1.0↗

You will find next a non-comprehensive list of datasets for research (or other purposes, like learning how to tackle large datasets, and so on).

Previous MSc students have (successfully) worked with the following secondary datasets:

Go back to questions


Do I need to fill out the Ethics forms?

First of all, a clarification of terminology:

  • Primary dataset refers to the first hand data gathered by the researcher himself.
  • Secondary dataset is data that a researcher has not collected or created themselves.

There are (at least) three situations where you must fill out the forms:

  1. You are generating a dataset involving humans, ie, containing human subject data.
  2. You are working with a dataset involving humans, idem.
  3. You would like to do usability testing with humans.

At any case, you must always check the dataset's license to see whether or not you can use it.

Doing research without proper ethical approval could result in disciplinary action, and even invalidate your research altogether, so please, take it very seriously.

Go back to questions


Can I use LLM/AI-generators (eg, ChatGPT)?

Yes, to help you (just as a calculator helps you), not to do the whole writing process for you.

If you get caught, it can have harmful consequences, just as any plagiarism.

These LLMs are helpful to help you clarify your ideas, and so on, however, the original thinking should be done by you.

Remember, it is YOU that will have to defend your ideas, so you might as well actually engage with your topic in a serious manner.

Go back to questions


How to prepare for meetings?

First and foremost, ask yourself:

  • "Is this meeting really necessary?"
  • "Can I try to work on my doubts before asking for a meeting?"

Only then ask for a meeting, either face-to-face or on-line (be especific on your request).

Send your availability by e-mail beforehand: select up to three day/time slots over the working week (Mon-Fri) during business hours (9am-5pm).

Time is a limited resource for anyone, so before the meeting, prepare by adhering to the following:

  • Be on time.
  • If you can't make it, notify before the meeting, not afterwards.
  • Have with you specific questions or send them by e-mail before the meeting.
  • Make notes during the meeting with action items, ie, things to do.
Go back to questions


Which AI/ML algorithm should I employ in my research?

Look at this self-explanatory image:



Go back to questions













Social media

Follow me on LinkedIn↗ or check out my ResearchGate profile↗.