Overview
At a first glance, data systems today are complex, with various components, tuning knobs and delicate design decisions.
However, at their core - as researchers and practitioners have been observing at least anecdotally - are simple ideas that work well in practice.
Simple ideas are easier to understand, less prone to bugs and more likely to be implemented by real products.
Keeping individual design points simple also helps reduce system complexity and improve maintainability.
However, this does not mean ignoring the necessary complexity or tricky corner cases.
Meanwhile, it also often takes a tremendous amount of experience to propose such ideas.
The goal of SiMoD (courtesy to Lawrence Benson for coming up with the name!) is to promote simple but non-trivial and effective ideas, by gathering researchers and practitioners to share their perspectives, new ideas, and experience on building practical data systems.
Workshop Chairs
Call for Papers
The Workshop on Simplicity in Management of Data (SiMoD) aims to promote simple but non-trivial and effective ideas.
SiMoD gathers researchers and practitioners to share their perspectives, new ideas, and experience on building practical data systems.
The workshop is open to basically all data systems topics, including but not limited to both classic database systems topics and data applications, such as
query processing, transaction processing, query optimization, storage management, networked and distributed database systems, data management in the cloud and self-driving database systems.
Submission Tracks
We invite two types of contributions: (1) novel ideas and (2) experience.
- Novel Ideas (up to 4 pages): Papers in this track should present early-stage, original ideas that were not proposed in the past.
The submission should be concise and distill the core idea on specific problems being solved, with reasonable evidence (e.g., preliminary experimental results) showing the idea's applicability.
We also encourage the authors to include the limitations of the proposed ideas.
Submissions in this track should not exceed four pages, including everything, such as references.
- Experience (1-page extended abstract or up to 10 pages): This track focuses on "war stories" and experience reports - existing ideas but with a focus on the process and results.
That is, how the idea came into its final/current form.
We especially encourage the submissions to reason about the possible tradeoffs that were made during the process.
For example, some surprisingly simple, elegant designs were in fact complex initially and/or came with many restrictions, but got "simplified" and refined through iterations.
The purpose of this track is mainly to encourage practitioners to discuss their experience;
the value lies more in the presentation itself and the paper will mostly serve archival purposes.
Submissions in this track are given much flexibility and can be (1) as short as a 1-page extended abstract or (2) as long as 10 pages to give the full picture, including everything.
Submission Instructions
Manuscripts should follow the latest ACM proceedings format (double-column) and be submitted as PDF files at:
https://cmt3.research.microsoft.com/SiMoD2024.
SiMoD is a single-anonymous workshop.
Authors are expected to include their names and affiliations on the first page of the manuscript.
Evaluation Criteria and Reviewing Process
Determining whether an idea is "simple but effective" is generally subjective and difficult (if not impossible) to quantify precisely.
When evaluating submissions, the PC will consider several helpful (although imperfect) metrics, including but not limited to:
- Whether the gist of the idea can be explained in few pages and does not require a long list of prerequisites.
- Whether empirical evaluation shows good performance results.
- Whether the amount of effort needed to implement the idea is small (e.g., as demonstrated by low lines of code); authors are encouraged to summarize the implementation and include source code in the submission.
As an example, the
B-link tree by Lehman and Yao is a classic that would fit the above evaluation criteria.
