Imagination in Computer Science Research

FSE New Faculty Symposium 2024
Imagination
in Research
All comments are for discussion only.
Abhik Roychoudhury
Department of Computer Science,
School of Computing,
National University of Singapore

FSE New Faculty Symposium 2024 2
Cultivates skills
Nurtures a mindset
Positive outcomes
However, there is a process
I…
I…
I…
(HIGH …)
(HIGHER …)
(HIGHEST)
Embrace and leverage the
process, while integrating
your work to achieve…
QUALITY!
RESEARCH?

THE THREE I’s
OF RESEARCH
3

RESEARCH
REQUIRES SKILLS
FOUNDATIONAL SKILLS
4
Assessing contributions in state-of-the-art papers
• Essential for paper reviews inside or outside program committees.
Comparing contributions across clusters of papers
• Chart the growth of a research area over time;
• Map evolution of a research field over time; and
• Provide guidance on future research directions.
Identifying emerging research trends in recent publications
• Evaluate potential feasibility.

ESSENTIAL SKILLS
RESEARCH
REQUIRES SKILLS
Requires effort, discipline and analytical skills
(particularly when analyzing papers and distilling
their core positions).
Involves proactive, analytical thinking beyond
mere effort — the first step in research.
5

THE FIRST I
Have Initiative &
(Genuine) Interest
Involves proactively seeking perspectives and knowledge
beyond your research group and doing so analytically.
6

KEY
CONSIDERATIONS
+ Choosing timely and impactful problems
• Choosing from known problem sets;
• Assessing importance and impact; and
• Identifying necessary technical skills and background.
Justifying why you are uniquely positioned to tackle
this problem at this moment
• Engaging in introspection and reflection
Have Initiative & (Genuine) Interest
7

THE SECOND I
Introspection
Involves looking within yourself.
8

KEY
CONSIDERATIONS
+ Critical reflections
• Identifying less apparent qualitative factors.
• Gaining clarity on the research's contribution.
• Moving beyond metrics like paper counts or citations.
• Advancing the field through impactful contributions.
• Emphasizing qualitative arguments over quantitative measures.
• Elevating the discourse to foster inclusivity and new contributions.
Introspection
9

THE THIRD I
Creating opportunities for others to contribute.
Extending beyond excellent problem-solving skills.
Having multiple ideas and the ability to assess
their contributions.
How do you decide which one to pursue?
Proposing a completely new direction?
10

KEY
CONSIDERATIONS
Utilitarian, Technical,
Conceptual Contributions
Focus is on the imagination such as invention of
program representations by Fran Allen
Invention of Unix operating system by Ken
Thompson, which has influenced Apple’s MacOS
The Third I
11

LOOKING BEYOND
TRENDING AREAS
12
Tenure takes five years
• What's hot in computer science in 2024 is unlikely to
remain so in 2029, as history shows.
Where do you envision yourself post-tenure?
• Avoid bandwagon jumping; "If you see a bandwagon,
it’s too late.”
Focus on conceptual contributions that chart
new directions
• Go beyond technical contributions; and
• Transform cool areas into hot ones, leveraging X.

THE THREE I’s
OF RESEARCH
Contributions 3 Is of Research
Utilitarian
Technical
Conceptual
Initiative & Interest
Introspection & Intuition
Imagination
Innovate in problem space & solution space
13

BUILDING THE
RIGHT TEAM
Investing substantial time in discussions to identify strengths
and weaknesses of students.
• Not to rush into writing papers at the first opportunity.
Feedback from my graduated students suggests higher
standards were achieved.
• Students sometimes disagree when I stress the importance of
preparing for a research career early on.
Encouraging conceptual contributions over just algorithmic
improvements.
• Finding the right level of abstraction.
Recognizing the need for more teaching opportunities —
students express this need.
• All PhD students are involved in tutoring.
• Ongoing developments in our setup in Asia.
Initially did not conduct targeted
recruitment.
Allowed students to self-select —
resulted in a slow start.
Over time, cultivated a highly
diverse research team with a wide
range of ideas.
14

…that the one who comes to teach
learns the keenest of lessons,
while those who come to learn learn nothing…
– J.M. Coetzee
“
15

SHARING MY
EXPERIENCE
Returned to Asia in January 2001 after completing my PhD. Three key observations during that time:
Achieve diversity in research teams
• Initially led to a slow start.
Beyond mastery of one technology
• Avoiding trends like MC (then) or DL (now);
• Emphasizing conceptual innovations alongside technical ones; and
• Exploring new usage scenarios for technologies beyond scalability.
Qualitative way of looking at research
• Steering away from quantitative emphasis.
• Adopting a thematic outlook rather than focusing solely on outputs.
• Encouraging a diversity of ideas within a thematic framework, rather than sticking to specific
technologies.
• Drawing from both formal (PhD) and informal (post-PhD) learnings.
Embracing risks in my journey
and encouraging my PhD students
to take risks in research.
16

THE
RIGHT TEAM
17
Cultivating imagination in the team
Sharing and convincing
On how novel a proposed idea is
Qualitative outlook instead of
focusing solely on results
Diversity
Not only explicit diversity
e.g. geographical
By a corollary, beyond
mastery of one technology
Several projects impossible
without this approach

(EXCITING)
TECHNICAL
CONTRIBUTIONS
Much of the work in SE is technical!
Does our technical work ignite others’ imagination?
Not just enhancing algorithms for efficiency or sustainability
• Technical contributions go beyond algorithmic improvements.
• Transplantation: Such as integrating temporal logics from philosophy into CS
to describe program properties.
• Cross-fertilization: Like applying algorithms for logical reasoning in temporal
property model checking.
18

TECH RESEARCH
SHOULD BE PRACTICAL
AND IMPACTFUL
Most research today is only used by other research groups.
Genuine concern among funders about the lack of
innovation in research being deployed.
Also concerns that prioritizing translation-oriented work
could compromise research quality.
Companies that have concerns will prefer reimplement solutions
19

RESEARCH IMPACT:
A FRESH LOOK
Adapted from National Research Foundation Singapore FRC report.
Diagram conceptualized and drawn by Abhik Roychoudhury, but in a separate capacity in NRF as Team Lead of NRF Foundational Research Capability study 2021.
20
Formulate general
research problem
State of the practice
Foundational
Research
Applied
Research
Translation
Concerns
about quality
More frequent
Quality
and impact
Less frequent
Roundtrip Engineering for growing mature R&D ecosystem

“EXCITING”
RESEARCH!
Discussions at NUS
for research planning
21

ACKNOWLEDGEMENT
DISCUSSIONS WITH NUS ALUMNI
… and many
others
Marcel Boehme
MPI, Germany
Sergey Mechtaev
UCL, UK > Peking University
SIGSOFT Outstanding
Dissertation
Van-Thuan Pham
University of Melbourne,
Australia
Shweta Shinde
Assistant Professor
ETH Zurich, Switzerland
Dawei Qi
Secidea, Shenzhen,
China
Shin Hwei Tan
Concordia University,
Canada
Xiang Gao
Beihang University,
China
Jooyong Yi
UNIST, South Korea
Sudipta
Chattopadhyay
SUTD, Singapore
Bruno Oliveira
Hong Kong University,
Hong Kong
Vivy Suhendra
NUS, Singapore
22

Thoughts?
abhik@comp.nus.edu.sg
Abhik Roychoudhury
Department of Computer Science,
School of Computing,
National University of Singapore

Imagination in Computer Science Research

More Related Content

Similar to Imagination in Computer Science Research

More from Abhik Roychoudhury

Recently uploaded

Imagination in Computer Science Research