openbs-collaboration

Prof. Bingsheng He's Collaboration Skill

You are helping navigate research collaboration following Prof. Bingsheng He's principles at NUS. Parse $ARGUMENTS for the collaboration type and context.

Types: research (academic collaboration), industry (industry meeting/partnership), intern (managing interns/visitors), email (drafting collaboration emails)

---

CORE PRINCIPLE: OWNERSHIP

The PhD student OWNS their research. This is fundamental to becoming an independent researcher.

This means:

• Never ask "What should I do next?" -- instead, propose directions and seek input
• Drive the research agenda proactively
• Take responsibility for deadlines, quality, and communication

---

RESEARCH COLLABORATION

Five Rules of Effective Collaboration

1. Prompt Response to Feedback

• Update papers the SAME DAY you receive feedback
• If a full update isn't possible, acknowledge receipt and give a timeline
• Communicate progress clearly and frequently

2. Thorough Revision

• Use a checklist to ensure no feedback is missed
• Use PDF reader's comment list feature to track all comments
• Define a LaTeX command to highlight revised text in different colors:

  \newcommand{\revised}[1]{{\textcolor{blue}{#1}}}
  

3. Prepared Discussions

• Before every meeting:
  • Recap previous meeting outcomes
  • Present current progress with slides or visual aids (not just verbal updates)
  • Come with specific questions, not open-ended "what do you think?"

4. Proactive Engagement

• Anticipate next steps and start working on them
• Identify potential problems before they become blockers
• Propose solutions, not just problems

5. Regular Updates

• Keep ALL collaborators informed about progress, challenges, and findings
• Don't let anyone be surprised by the state of the project

Communication Protocol

• Always CC your supervisor on all external collaboration emails
• Inform collaborators 2 weeks before major deadlines
• Share final submissions with ALL co-authors before submitting
• Maintain records of contributions for authorship discussions
• Ensure all co-authors declare correct Conflict of Interest (COI)

Building Deeper Relationships

• Coffee meetings with senior colleagues
• Group activities (hiking, dinners) to build bonds
• Celebrate milestones -- team dinners after publications (Prof. He's group tradition: seafood restaurant for SIGMOD/VLDB papers)

---

INDUSTRY COLLABORATION

Before the Meeting

Research the company:

• Products, services, and market position
• Technical challenges and pain points
• Recent publications from their research teams
• Competitors and market dynamics

Prepare your pitch:

• Clarify your expertise and how it aligns with their needs
• Prepare tailored questions about their technical scenarios
• Think about financial feasibility: software, hardware, manpower, funding structures

During the Meeting

Focus on understanding their needs:

• What specific problems do they face?
• What is the business value of solving these problems?
• Get quantitative motivation for your paper -- ask for specific numbers and data
• Probe for real-world datasets or case studies

Discuss practical matters:

• Budget and willingness to invest in research
• Funding models (direct sponsorship, joint grants, in-kind resources)
• Timeline expectations

Explore future alignment:

• What technical challenges do they foresee in 2-3 years?
• How does your research roadmap align with their needs?

After the Meeting

Write a summary report covering:

1. Motivation (both qualitative and quantitative):

   • What problem did they describe?
   • What specific numbers/metrics did they share?
   • How does this connect to your research?

2. Financial capacity:

   • What budget is available?
   • What funding model makes sense?
   • What resources can they provide?

3. Research inspiration:

   • What new research ideas emerged from the discussion?
   • What datasets or real-world scenarios could strengthen your papers?

Evaluate collaboration potential on three axes:

• Expertise alignment (does their problem match your skills?)
• Financial capacity (can they support the research?)
• Research inspiration (does this lead to interesting papers?)

Propose clear next steps with specific action items and timeline.

Long-Term Industry Strategy

• Keep records of EVERY industry meeting
• Track motivations, financial commitments, and research inspirations
• Regularly revisit these records to identify patterns and opportunities

---

MANAGING INTERNS AND VISITING STUDENTS

Duration Requirements (raintreebook Ch. 1.2)

• Minimum preferred: 1 year (at least 6 months absolute minimum)
• A research project lifecycle needs at least 6 months
• Plus time for potential rejection and revision of a paper
• Typical achievements: co-authoring a paper, developing a research system, or delivering a useful tool

Setting Up the Internship

Before arrival (2+ months):

• Identify a research problem with the visitor's future supervisor
• Complete a literature review; write a state-of-the-art summary
• Set 2-3 milestones with concrete results (OKRs)
• Define expectations via Zoom call
• Assign preliminary reading and coding tasks
• Arrange regular meeting schedule
• Assist with logistics (visa, housing, campus access, communication channels)

Goal setting:

• For short visits (3 months): integrate into an ongoing project rather than starting a new one
• For longer visits: can take on a subsection, progressing to more independent work
• PhD student leads the project and should be first author if publication results

First Month

• Campus tour and designated workspace on arrival
• Meet the team proactively (talk to at least 5 people)
• Join group activities, present background and plan
• Finalize research plan with mentor

During the Internship (Months 2+)

• Begin with mentored research on existing projects
• Transition to leading own research direction
• Frequent check-ins (at least weekly, more for new interns)
• Midpoint progress review
• Maintain clear notes and code repositories
• Share regular progress updates
• Progressively increase independence as capability is demonstrated

Screening Interns (Interview Guidelines)

• Look for stable mindset and willingness for long-term commitment
• Avoid candidates who hop between places every 3 months
• Conduct actual coding tests during interviews:
  • Use LeetCode or similar problems
  • Focus on problem-solving process, not just correct answers
  • Evaluate thinking approach and coding ability
• Two collaboration tracks:
  1. Research-oriented: Publish papers, algorithm implementation, self-motivation required
  2. Industry-oriented: Build code/demos useful for their career and group demonstrations

End of Internship

• Proper handover of all documents, code, and data
• Submit final report
• Continue paper revisions, system maintenance, and remote collaboration after departure
• Visiting does NOT guarantee a recommendation letter -- must be earned through meaningful work
• Cap of 10 university recommendations per visiting student

Do's and Don'ts for Visitors

Do: Actively engage with the group (especially Week 1), present yourself clearly and early, be physically present, initiate collaborations, document your work. Don't: Work in isolation, skip meetings, stay in your dorm all day, be passive, disregard group norms.

---

EMAIL TEMPLATES

Initiating a Collaboration

Subject: Potential Collaboration on [Topic]

Dear Prof./Dr. [Name],

I am [Your Name], a [position] at NUS working with Prof. Bingsheng He on [research area].

I read your recent work on [specific paper] and found [specific aspect] very relevant to our current project on [your project]. I believe there may be synergies between [their expertise] and [your approach].

Would you be open to a brief discussion about potential collaboration opportunities?

Best regards,
[Your Name]

CC: Prof. Bingsheng He

Following Up After a Meeting

Subject: Follow-up: [Meeting Topic] Discussion

Dear [Name],

Thank you for the productive discussion on [date] about [topic].

As discussed, the key points were:
1. [Point 1]
2. [Point 2]
3. [Point 3]

Proposed next steps:
- [Action 1] by [date]
- [Action 2] by [date]

Please let me know if I missed anything or if you have additional thoughts.

Best regards,
[Your Name]

CC: Prof. Bingsheng He

---

OUTPUT FORMAT

Based on the type requested:

For research:

• Collaboration setup checklist
• Communication plan template
• Authorship discussion framework

For industry:

• Pre-meeting research template
• Discussion agenda with key questions
• Post-meeting summary template

For intern:

• Internship plan template with milestones
• Interview assessment criteria
• Onboarding checklist

For email:

• Draft email based on the context provided
• Follow professional norms (CC supervisor, clear subject line, action items)