The 1st International Workshop on LiveWriting: (Experiments in Research Collaboration)

Stop falling asleep listening to other people’s talks. Come to a meeting, leave with new ideas. Show up with a vague notion. Listen to everyone else’s notions. Form affinity groups. Work together to refine your ideas into an action plan that is submitted to a journal (Automated Software Engineering)

A new kind of workshop:

• We meet together to work together. At this meeting, everyone collaborates to do fresh research.
• We don’t sit and watch talks. Instead, we group together to write papers. Arrive at 9 am with a short abstract; leave at 5 pm as a co-author on a finished paper. No homework, before or after.
• Teams, not solos. Abstracts are pre-clustered into writing teams, matched on who you’ll write well with.
• Fast-track to journal. Papers clearing the editor’s bar are invited to a special issue with expedited review. Real venue, real provenance.
• Proof-of-Presence authorship. To maintain academic integrity and satisfy the strict authorship requirements of the Automated Software Engineering journal, the workshop employs a tracked Proof-of-Presence framework. Utilizing shared Overleaf repositories and version-controlled Git history, each participant’s distinct text contributions, prompt designs, and structural edits will be transparently logged. Journal representatives will oversee this digital ledger during the revision squads, ensuring that co-authorship is directly tied to active, documented collaboration on the day of the event.

Why this approach?

We need a new way to do research. Somewhere along the line, writing the paper became the job — the thing we sink months into, alone in a back room. The research paper should be the least of our efforts, not the most.

To fix that, we will leverage Large Language Models to accelerate syntactic mechanics—such as initial prose expansion, structural formatting, and preliminary literature mapping. This optimizes the workflow, allowing human researchers to dedicate their cognitive bandwidth to high-level conceptual modeling, critical evaluation of AI outputs, rigorous argument validation, and cross-disciplinary synthesis

So let’s see if that works. Let’s create writing workshops where the sum is greater than the parts, where researchers meet to craft their papers, and where everyone collaborates to do fresh research. Here’s the plan:

The day runs:

• 09:00 — teams form around clustered abstracts
• 09:40 — morning writing workshops (results, questions, scope)
• 12:00 — round-robin panel across teams
• 13:00 — drafting sprint (LLMs welcome as tools)
• 14:00 — complete drafts
• 14:00 – 16:30 — revision polish in Overleaf
• 16:30 — papers 95% done

Three weeks later: submitted to the ASE journal.

Call for One-Page Abstracts

Submit PDFs to https://livewrite26.hotcrp.com/.

Submission Guidelines

• Format: 1 page max for text, 1 page for references, ACM sigconf two-column.
• Content: Your name and email at the top, an eye-catching title, and headings to describe the BASE:
  • B - Bet (Motivation): The core “why” and a sharp, open question.
  • A - Approach (Methods): The proposed innovation/ new pint of view. The "how of the proposal immediate tools, code, or accessible materials.
  • S - Sources (Data): The “what”—available data, case studies, or delta data to be collected.
  • “Sources” is required for registered report proposal, but is optional for position paper.
  • E - Experience (People): The “who”—prior results, relevant publications, and Google Scholar links to verify expertise. Include yourself in this list.

Team Creation

Accepted abstracts will be shared with the authors of other accepted abstracts to form birds-of-a-feather affinity groups. Abstracts may be submitted by individuals or groups. After submission, accepted authors will be asked for their preferences for the groups they want to work in.

• While we aim to respect those groupings, the chairs reserve the right to tinker with the groupings to (e.g.) ensure that groups are not overwhelmed by research newcomers or dinosaurs.

Paper Types

We ask authors to write a one-page abstract that proposes either of the following:

Type 1: Proposal to Complete a Position Statement: (in one day, writing in a group)

Such statements present a clear, well-argued stance on an open problem, emerging trend, or controversial issue in the field. Rather than reporting mature results, they articulate a viewpoint, provoke discussion, or propose a research agenda. Preliminary evidence, early experience, or thought experiments are welcome, but the emphasis is on the strength and novelty of the argument.

Modulo paper quality, such position papers might become one publication in the ASE journal

Type 2: Proposal to complete to write a Registered report, part1 (in one day, writing in a group)

Registered reports split review into two stages.

Authors first submit their motivation, hypotheses, and analysis plan, but no results. Plans that pass review earn in-principle acceptance, guaranteeing publication of the finished work whatever the outcome, so long as authors follow the registered protocol. The completed paper is then reviewed for adherence to the plan, not for the appeal of its findings.

Modulo paper quality, such position papers might become two publications in the ASE journal

• Paper one would be accepted just after the workshop
• Paper two would be due 4 months later

For this type of submission, abstracts should propose a 4-month study capable of completion by 2-4 people (standard university-based researcher study). Such a study should be a few months long, with work interleaved with other commitments such as teaching, service, grant writing, and other research projects.

Topics

LIVEWRITE welcomes submissions addressing topics across the full spectrum of Automated Software Engineering. Topics of interest include the following and are grouped into the following nine research areas. Please note that these topics are by no means exhaustive– if you think your topic is relevant to ASE, please check with the LIVEWRITE chairs:

Requirements and Design:

• Requirements elicitation and management, traceability analysis
• Software architecture and design
• Modeling and model-driven engineering
• Software product lines
• Component-based or service-oriented systems
• Object-oriented or aspect-oriented systems

Testing and Analysis:

• Regression, mutation, and model-based testing
• System, unit, and integration testing
• Black-, grey-, and white-box fuzzing
• Automated program repair and synthesis
• Static and dynamic analysis
• Empirical program analysis

Maintenance and Evolution:

• Debugging and fault localization
• Refactoring and reengineering
• Reverse engineering
• Software reuse
• API design and management

Human and Social Aspects:

• Software engineering processes (e.g., agile, DevOps)
• Green and sustainable technologies
• Software engineering ethics and values
• Software economics
• Systematic code review and inspection
• Program comprehension and visualization
• Crowd-based and collaborative software engineering
• Human-computer interface

AI and Software Engineering:

• Autonomous and self-adapting systems
• Search-based software engineering
• Recommender systems
• AI4SE
• SE4AI

Software Analytics:

• Mining software repositories
• Analysis of mobile apps, app stores
• Data-driven user experience understanding and improvement
• Data-driven decision making in software engineering
• Software metrics (and measurements)

Formal Aspects of Software Engineering:

• Formal methods and model checking
• Programming languages
• Domain-specific or specification languages
• Software validation and verification

Security and Other Non-Functional Properties:

• Security and privacy
• Dependability and safety
• Reliability and availability
• Performance

Would you prefer to work with other researchers to create new research rather than passively listen to talks? Do you want to help put work back into workshops? LiveWrite has you covered!

LiveWrite is an experiment in making again workshops the kind of interactive, science-focused experience they were always meant to be. We invite you to participate in the workshop. The format is as follows:

• Based on the first submission round, some groups are already forming. If you like, you can join our Discord immediately or just show up right at the workshop at ASE in Munich.

• At the workshop, we will have intensive discussions and joint technical work to produce joint papers. You are welcome to join any group with shared interests.

The plan is to do most of the work right there, right then (commitments are expected to be full-day). In particular, two types of papers are expected:

• Vision/Position papers: jointly, you discuss a current challenge in automated software engineering in detail and produce a joint paper.

• Registered reports: you create a registered report, which describes motivation, hypotheses, and an analysis plan, but no results. This is particularly interesting if you already have some studies in mind and are looking for collaborators. This is the first published paper and can be the basis for after-the-workshop work, which will then lead to a full paper.

Guiding principles are:

• Fast-track to journal. Papers clearing the editor’s bar are invited to a special issue with expedited review. Real venue, real provenance.

• Proof-of-Presence authorship. To maintain academic integrity and satisfy the strict authorship requirements of the Automated Software Engineering journal, the workshop employs a tracked Proof-of-Presence framework. Utilizing shared Overleaf repositories and version-controlled Git history, each participant’s distinct text contributions, prompt designs, and structural edits will be transparently logged. Journal representatives will oversee this digital ledger during the revision squads, ensuring that co-authorship is directly tied to active, documented collaboration on the day of the event.

• Regrouping and new groups allowed. While initial abstracts guide ideas and already now lead to initial groups, regrouping and the formation of new groups based on shared interests are possible at any time.

• Generative AI is a valid partner for your work. (But teams still need to include at least two humans from two different organizations.) The organization team provides a toolbox to inspire how you use LLMs in your work. We also invite you, but do not require you, to contribute to a survey about your experience using LLMs in your scientific work.

Ok, you got me curious: what should I do?

1. Join the Discord.

2. Check out our toolbox.

3. Check out our guidelines.