Art Of Research
John H. Reif
Research in general is something that you must have passion for, and truly love.
Leaving Research Areas:
á If you find that eventually you do not have a passion for a particular topic of research, look broadly and find another.
á Even if you are a leading expert with lots of papers on the topic, you need to find the courage to walk away from your research topic if you find your passion is no longer in that topic, or you feel that there are better opportunities in other topics.
á Only do this change of research topics with some deliberations, and do it gradually, allowing for completion of grants, prior papers and completion of studentÕs degrees.
Finding a New Research Area:
á Choose a research area that is ÒhotÓ, with a lot of new progress, at least adequate existing funding, at least a few good people beginning to enter the area.
á It is much easier to have impact on a new emerging research area rather than a well-established research area.
á Try to make sure you have some unique talents and skills that might be of use in the new topic.
To initiate Research in a particular area, consider both:
á the broad scope of the research, as well as
á the narrow topic you are intending to work on.
Learn everything you can on the research topic (particularly the narrow) by:
á literature search (e.g., Google scholar),
á library books,
á Amazon (for more recent books).
Read carefully all these papers and books and master their techniques:
á Start with a carefully chosen set of introductory papers and books, then to the more technical papers and books.
á If there is homework in any book, solve every homework.
Know the terminology, methods, tricks of all work related to your intended research.
Form a Reading Group with other potential collaborators
á Meet regularly on a fixed schedule.
á Each week, one member of the group should present other prior work.
á Discuss potential research projects after each presentation.
Keep a Research Notebook with
(2) Technical Techniques or Tricks of the Trade: these may be implicit in papers so keep an eye out for them. They also might be explicitly described in good text books.
(3) Open Problems: these can often found at the end of papers in conclusion, or end of talks.
o Even if you do not solve these,
o there may be related problems you can solve,
o that that may be even more important.
(4) Your Ideas: whatever comes to mind.
á In every field, there are key people that generate many important ideas and breakthroughs.
á Know absolutely everything done by the key people, and master their techniques.
á Try pretending you are a particular such key person and try writing a paper pretending you are them and now writing their next great paper.
á Also, go to the most important conference on the research topic and great everyone there, both the key people as well as everyone else, become their friend and consider them as potential collaborators - value them.
Initiating a Research Project Leading to a Paper:
á Start with a problem that seems important.
á You will need some novel research idea.
Coming up with a Novel Research Idea:
á A novel research idea may not come right away, so be patient.
á A novel research idea appear suddenly, in a dream or with a tingle down the back of your back.
á So value your dreams and intuition, they are your subconscious giving you gifts of novel ideas.
Refining You Research Idea and Conducting the Research
á Draw up a 1 page Summary of the research paper you intend to write.
á Determine the key tasks to be done and milestones to achieve
á Draw up a very brief work plan and schedule of research to be done.
Find Collaborators and Research Team Members for Your Research Project
á List the key techniques and skills needed in your research.
á If you are missing any, locate potential collaborators in the project.
á It helps if you are already friends and trust that potential collaborator. If not, make this so.
á Make sure you have discussed what you expect from your collaborator, and roughly what time table is.
á Keep in regular contact with your collaborator, with regular Skpe meetings if not local.
by consensus the research team leader.
Be Flexible on your Research Project Direction & Goals
á You may not have progress on key tasks.
á Be flexible on time duration and goals – it is OK to revise these to get around emerging difficulties.
á You may also be able to later determine further new objectives and potential results.
á Reward yourself and/or collaborators when a milestone is achieved, with a gift or special meal (even if it just you doing the research).
Paper Drafts and Revisions
á Always give a date for the paper draft.
á Have someone in your research team (or yourself) be the responsible person for the current draft and distributing via email or dropbox.
á Coauthors need to be given guidance on what sections they are responsible for edits or expanding.
á Decide with consensus on list of paperÕs authors and order of authors.
á Never include anyone that did not contribute to the paper in a significant way.
á Never exclude anyone contributed to the paper in a significant way.
á Decide on the conference or journal the paper is to be submitted with care, based on:
o fit with you research paper topic,
o visibility, citation index,
o difficulty to get accepted.
á Submit to the highest quality conference or journal the paper has some likelihood of acceptance:
o rejection is not so bad, since you should receive detailed referee comments.
á If rejected, edit the paper,
o addressing all referee comments, even if you feel they are unreasonable (except in particular circumstances where the referee comments are clearly incorrect).
á If resubmitting, then write a cover letter explaining how all referee comments are addressed.
á If instead you submit to another journal or conference, choose:
o one where you have a bit more likelihood of acceptance,
o but not too low on ratings.
Doing Research is not easy: it is often very demanding but it gives wonderful rewards, the most important are not external but internal.
At the End:
Research in general is something that you must have passion for, and truly love.
Worlds Highest Impact Journals:
á Proceedings of the National Academy of Science(PNAS):
á Science, Nature, and PNAS review papers only in topics considered of high importance in all of science.
á The acceptance rate of Science, Nature, and PNAS is very low, approximately 5%; most papers submitted are immediately rejected out of hand, without review.
á Note: The members of the National Academy of Science can submit one paper a year to PNAS, so this is a back door for consideration, but even so the paper is still very rigorously reviewed.
Citation Numbers: The citation numbers for Science, Nature, and PNAS papers are often very high, and can often number to the hundreds.
á It is not essential to publish in Science, Nature, and PNAS to be a highly cited computer scientist.
á Most very well known computer scientists have no papers in any of Science, Nature, or PNAS, but still had important contributions to computer science and are highly cited.
á Any paper from a computer scientist published in Science, Nature, or PNAS can significantly increase the visibility of the authorÕs Department, University, and the overall field of Computer Science.
ReifÕs papers in Highest Impact Journals:
Science: contributed papers
 Yan,H., Park, S.H., Finkelstein, G., Reif,JH., & LaBean,TH., DNA-Templated Self-Assembly of Protein Arrays and Highly Conductive Nanowires, 2003, Science, Vol. 301, pp. 1882-1884, PubMed PMID: 14512621.
 Yin,P. Hariadi,RF., Sahu,S., Choi,HMT, Park,SH., LaBean,TH., Reif,JH., Programming DNA Tube Circumferences, 2008, Science, Vol. 321. no. 5890, pp. 824–826 DOI: 10.1126/science.1157312.
Science (Perspectives) – these are invited review papers only
 Reif,JH., DNA Computation - Perspectives: Successes and Challenges, Science, 296: 478-479, April 19, 2002. PubMed PMID: 11964464.
 Reif,JH., Perspectives: Biochemistry: Scaling Up DNA Computation, 2011, Science, Vol. 332, pp. 1156-1157 DOI: 10.1126/science.1208068
Nature: contributed papers
 Mao,C., LaBean,TH., Reif,JH., Seeman,NC., Logical Computation Using Algorithmic Self-Assembly of DNA Triple-Crossover Molecules, 2000, Nature, vol. 407, pp. 493-495 C. Erratum: Nature 408, 750-750.
Proceedings of the National Academy of Science (PNAS):
 Liu,D. Reif,JH., LaBean,TH., DNA nanotubes self-assembled from triple-crossover tiles as templates for conductive nanowires, 2004, PNAS, Vol. 101, No. 3, pp. 717-722, PubMed PMID: 14709674.
 Yan,H., LaBean,TH., Feng,L., & Reif,JH., Directed Nucleation Assembly of Barcode Patterned DNA Lattices, 2003, PNAS, Volume 100, No. 14, pp. 8103-8108, PubMed PMID: 12821776.