If you are a graduate student trying to publish, check this out!
The goal of this post is to share some guidelines with graduate students on publishing in journals and conferences. It is based on my past experience as a PhD student. One of my AI (deep learning) papers [1] recently crossed 1000 citations per Google Scholar [2]. According to Web of Science™ database [3], only ~ 0.026% papers have over 1k citations. While it is great to see the post-publication impact the paper has had, the point up to the paper getting accepted was a lot of trial and error. This paper was rejected multiple times, took a couple of years, and underwent multiple iterations before it was accepted. I have tried to distill my learnings here into a guided process. Hopefully, the guidelines below will help you in your publishing journey.
Publishing is hard, especially if your school requires submission to high impact journals/conferences. Most PhD programs have publishing requirements, and it can even be the reason for holding up your graduation.
Goals
- Reduce time to publication
- Amplify impact of paper
Here are some of my learnings that you might find useful to reduce stress that comes with publishing. To draw an analogy, I think starting a startup company and the PhD publication process have a few things in common. Both have an uncertain future initially, but both could result in a novel contribution in their respective areas in the end. Just like a startup, your publication process needs a survey, vision, strategy, iterations, and scaling. My thoughts are organized below in the same sequence.
Survey
Read up on current state-of-the-art in the area you are trying to publish. Remember you are trying to add to what is already out there. Many publications are open-access, so finding papers on the latest research should not be an issue. Plus, your university should also be able to provide you with necessary access. Google Scholar and ResearchGate are great open sources. Given that technology and scientific research are advancing so fast, I would also follow top researchers and companies in your area of research in LinkedIn, Twitter, and similar sites for the latest updates. If applicable, I would also explore statistics and data repositories for exploratory analysis. On the pace of the survey, recommended numbers vary from reading 1 to 7 papers per week in preparation for literature review and having 30 to 200 papers included/referenced in your survey.
Vision
After the survey, develop a vision of what topic you want to publish in. A startup starts with a problem it intends to solve. The founders develop a vision that aligns with them and the problem they intend to solve. To help find a match for yourself, start with the following questions:
- What are some of the unresolved problems in your area of study/major? E.g., this [4] is a good reference list for unresolved problems in various disciplines.
- What were some of the active research areas discussed in the survey papers?
- Which course[s] excited you the most during your pre-qualifying exam coursework?
- What are the areas of expertise of your advisor?
Try to pick the topic from a region that has at least three overlaps in the Venn diagram above. Write an abstract and get feedback from your advisor. Ensure it includes your primary goal and proposed outline of the paper. The more critical the feedback, the more adjustments that will be required at this stage.
Strategy
During the initial stages of a startup company, its primary goal is to experiment, maximize learning, and land on building something novel the market actually needs, i.e., finding the elusive product-market fit. In your case, that would be landing on a publication, i.e., paper-publication fit. Both require strategy.
Before you come up with a strategy, consider the following:
- How much time and energy can you commit to publication effort? How are you going to balance other commitments?
- How much time does your advisor have to guide you? Try to see if you can set up a recurring touchbase with your advisor.
- What is your target time to get necessary publications to graduate? Ensure this is realistic.
- How are you managing stress? Note that ~ 50% PhD candidates in North America drop out before they get their degree [5]. You will need some type of life-hack to get through PhD.
Once your figure these questions out and do some retrospective, come up with a publication strategy that includes:
- Paper topic/problem
- Paper type, outline and target length
- Target list of publications to submit to:
– Rank by acceptance-difficulty/impact-factor
– Include their turn around (review cycle) times
– Ensure the publication meets all criteria set by your institution - Anticipated target date to achieve publication success
– account for n iterations and time to revise paper n times
Iteration and Pivot
Once you and your advisor are aligned on the strategy/focus, start your research and try to get to a point where you have something to show for (aka an MVP or minimal viable product in startup terms). Once you have an MVP paper, submit to journals/conferences starting at the higher difficulty range but have known shorter review cycles. Shorter review cycles let you iterate and incorporate feedback quicker and pivot strategy sooner, if needed.
Don’t get disheartened by rejection and instead use the critical (or harsh) feedback to make bold changes to your paper and/or strategy. Ensure that you are evaluating, responding to, and incorporating the feedback/gaps provided by the reviewers. Rejection with feedback is a blessing and a vital part of the process. Think of it as training an AI model. The weights of the artificial neural networks are adjusted based on feedback from labeled/training data during the training process. As shown in figure 3 below, these adjustments are initially higher and gradually tapers off when you get closer to paper-publication fit (or global minima in AI/ML).
Pick journals/conferences that align with your papers research and read the submission guidelines carefully. There is nothing worse than waiting for months for feedback, only to learn the paper you submitted can not be considered because it isn’t aligned with the journal/conference’s theme or you missed an important submission step. If applicable, it will help to research past papers from journal/conference you are submitting to and cite relevant work from those papers. Also, if your co-authors (e.g., advisor) have had successful papers in certain journals, try them as well.
Scale
You will gradually get the hang of the process that leads you closer to publication after a few iterations, as the rejection reasons get less critical. It could take a really long time to get there. Note that your first publication could most likely be the most difficult and take the longest time.
In machine learning, there is a technique called transfer learning [6], where you can apply knowledge gained from solving one task to another related task, with decreasing amount of learning effort. Just like that, all the validated learning from your first publication will come in very handy in speeding up your 2nd, 3rd and subsequent publications.
Faster feedback cycles and iterations are the key. Use that to your advantage to get to the necessary number of publications. Stick to what works for you, and continue to learn and tweak the process for more impactful publications. Good luck!
If you are a graduate student trying to publish, check this out!
The goal of this post is to share some guidelines with graduate students on publishing in journals and conferences. It is based on my past experience as a PhD student. One of my AI (deep learning) papers [1] recently crossed 1000 citations per Google Scholar [2]. According to Web of Science™ database [3], only ~ 0.026% papers have over 1k citations. While it is great to see the post-publication impact the paper has had, the point up to the paper getting accepted was a lot of trial and error. This paper was rejected multiple times, took a couple of years, and underwent multiple iterations before it was accepted. I have tried to distill my learnings here into a guided process. Hopefully, the guidelines below will help you in your publishing journey.
Publishing is hard, especially if your school requires submission to high impact journals/conferences. Most PhD programs have publishing requirements, and it can even be the reason for holding up your graduation.
Goals
- Reduce time to publication
- Amplify impact of paper
Here are some of my learnings that you might find useful to reduce stress that comes with publishing. To draw an analogy, I think starting a startup company and the PhD publication process have a few things in common. Both have an uncertain future initially, but both could result in a novel contribution in their respective areas in the end. Just like a startup, your publication process needs a survey, vision, strategy, iterations, and scaling. My thoughts are organized below in the same sequence.
Survey
Read up on current state-of-the-art in the area you are trying to publish. Remember you are trying to add to what is already out there. Many publications are open-access, so finding papers on the latest research should not be an issue. Plus, your university should also be able to provide you with necessary access. Google Scholar and ResearchGate are great open sources. Given that technology and scientific research are advancing so fast, I would also follow top researchers and companies in your area of research in LinkedIn, Twitter, and similar sites for the latest updates. If applicable, I would also explore statistics and data repositories for exploratory analysis. On the pace of the survey, recommended numbers vary from reading 1 to 7 papers per week in preparation for literature review and having 30 to 200 papers included/referenced in your survey.
Vision
After the survey, develop a vision of what topic you want to publish in. A startup starts with a problem it intends to solve. The founders develop a vision that aligns with them and the problem they intend to solve. To help find a match for yourself, start with the following questions:
- What are some of the unresolved problems in your area of study/major? E.g., this [4] is a good reference list for unresolved problems in various disciplines.
- What were some of the active research areas discussed in the survey papers?
- Which course[s] excited you the most during your pre-qualifying exam coursework?
- What are the areas of expertise of your advisor?
Try to pick the topic from a region that has at least three overlaps in the Venn diagram above. Write an abstract and get feedback from your advisor. Ensure it includes your primary goal and proposed outline of the paper. The more critical the feedback, the more adjustments that will be required at this stage.
Strategy
During the initial stages of a startup company, its primary goal is to experiment, maximize learning, and land on building something novel the market actually needs, i.e., finding the elusive product-market fit. In your case, that would be landing on a publication, i.e., paper-publication fit. Both require strategy.
Before you come up with a strategy, consider the following:
- How much time and energy can you commit to publication effort? How are you going to balance other commitments?
- How much time does your advisor have to guide you? Try to see if you can set up a recurring touchbase with your advisor.
- What is your target time to get necessary publications to graduate? Ensure this is realistic.
- How are you managing stress? Note that ~ 50% PhD candidates in North America drop out before they get their degree [5]. You will need some type of life-hack to get through PhD.
Once your figure these questions out and do some retrospective, come up with a publication strategy that includes:
- Paper topic/problem
- Paper type, outline and target length
- Target list of publications to submit to:
– Rank by acceptance-difficulty/impact-factor
– Include their turn around (review cycle) times
– Ensure the publication meets all criteria set by your institution - Anticipated target date to achieve publication success
– account for n iterations and time to revise paper n times
Iteration and Pivot
Once you and your advisor are aligned on the strategy/focus, start your research and try to get to a point where you have something to show for (aka an MVP or minimal viable product in startup terms). Once you have an MVP paper, submit to journals/conferences starting at the higher difficulty range but have known shorter review cycles. Shorter review cycles let you iterate and incorporate feedback quicker and pivot strategy sooner, if needed.
Don’t get disheartened by rejection and instead use the critical (or harsh) feedback to make bold changes to your paper and/or strategy. Ensure that you are evaluating, responding to, and incorporating the feedback/gaps provided by the reviewers. Rejection with feedback is a blessing and a vital part of the process. Think of it as training an AI model. The weights of the artificial neural networks are adjusted based on feedback from labeled/training data during the training process. As shown in figure 3 below, these adjustments are initially higher and gradually tapers off when you get closer to paper-publication fit (or global minima in AI/ML).
Pick journals/conferences that align with your papers research and read the submission guidelines carefully. There is nothing worse than waiting for months for feedback, only to learn the paper you submitted can not be considered because it isn’t aligned with the journal/conference’s theme or you missed an important submission step. If applicable, it will help to research past papers from journal/conference you are submitting to and cite relevant work from those papers. Also, if your co-authors (e.g., advisor) have had successful papers in certain journals, try them as well.
Scale
You will gradually get the hang of the process that leads you closer to publication after a few iterations, as the rejection reasons get less critical. It could take a really long time to get there. Note that your first publication could most likely be the most difficult and take the longest time.
In machine learning, there is a technique called transfer learning [6], where you can apply knowledge gained from solving one task to another related task, with decreasing amount of learning effort. Just like that, all the validated learning from your first publication will come in very handy in speeding up your 2nd, 3rd and subsequent publications.
Faster feedback cycles and iterations are the key. Use that to your advantage to get to the necessary number of publications. Stick to what works for you, and continue to learn and tweak the process for more impactful publications. Good luck!