Best Recommender Systems Workshop: Top Picks Compared

By MLRec Staff · Updated on September 24, 2026

The best recommender systems workshop options come down to venue prestige, review rigor, archival status, and travel cost, with roughly a dozen recurring, well-established venues each year. ACM RecSys alone hosts multiple half-day and full-day workshops, while SIAM’s MLRec workshop anchors the data-mining side of the calendar. Workshop papers are typically shorter (4-8 pages), reviewed by smaller committees, and sometimes not archived.

Key Takeaways

Two ecosystems dominate: ACM RecSys workshops (applications, evaluation, human-centered design) and SIAM/MLRec-style data-mining workshops (algorithms, scalability, theory).
Archival status is the single biggest differentiator. Some workshops publish proceedings in ACM DL or CEUR-WS; others are non-archival by design, which matters for how you cite the work later.
Review rigor varies widely. Invited-only, abstract-only, and full-paper tracks coexist; check the call for papers before assuming a workshop is “easy.”
Fit beats prestige. A specialized workshop (e.g., OARS for offline evaluation, HCREC for human-centered recommendation) can advance a paper more than a general venue.
Deadlines cluster in spring and early summer for fall conferences — plan a 3–6 month lead time for a full-paper submission.

Why Workshop Selection Matters More Than It Used To

Research on recommender systems has matured to the point where workshops are no longer a consolation prize for rejected conference papers. A workshop paper on offline assessment, equity, or conversational recommendation can reach a smaller, more expert audience than a general conference talk, and the discussion format—often 20-minute slots with 10 minutes of discussion—produces feedback that shapes a subsequent full paper. Researchers who view workshops as a strategic venue and not a fallback tend to get more citations and longer-lasting collaborations.

The tradeoff, however, is real. Workshop papers are typically shorter (4-8 pages), reviewed by smaller committees, and sometimes not archived. An unarchived workshop paper cannot be cited in the same way as a conference paper, and some hiring and promotion committees discard it. The right choice depends on whether you want visibility, feedback, or a quotable record – and these goals go to different places.

The Two Main Workshop Ecosystems

ACM RecSys Workshops

The ACM Conference on Recommender Systems (RecSys) is the flagship venue for the field, and its workshop program is the largest single cluster of recommender-systems workshops. Recent editions have hosted workshops spanning conversational recommendation, fairness and bias, human-centered recommendation, offline evaluation, and industry deployment. The RecSys workshops page is the canonical index; each workshop maintains its own site with its own call for papers, deadlines, and proceedings policy.

Strengths: proximity to the main conference (you attend both), high-quality attendees, strong industry participation, and often ACM DL or CEUR-WS proceedings.

Weaknesses: competitive acceptance, tight timelines, and a schedule that can conflict with main-conference sessions you also want to attend.

Related: — Browser-based, hands-on ML and data-science tracks you can start in 10 minutes.

SIAM Data Mining and MLRec-Style Workshops

The SIAM International Conference on Data Mining (SDM) hosts workshops oriented toward the algorithmic and statistical side of recommendation — matrix factorization, graph-based methods, scalability, and evaluation methodology. MLRec, the Workshop on Machine Learning and Data Mining for Recommender Systems, sits in this tradition: it attracts researchers who care about the underlying learning problem as much as the application.

Strengths: deeper methodological focus, smaller and more discussion-friendly, strong fit for theory and large-scale systems work.

Weaknesses: less industry visibility than RecSys, and a smaller pool of potential collaborators if your goal is deployment.

Our pick: — University- and industry-branded ML specializations with graded assignments and shareable certificates.

Comparison Table: How the Leading Workshops Differ

Workshop / Family	Host Conference	Typical Focus	Proceedings	Best For
RecSys workshops (general)	ACM RecSys	Applications, industry, HCI	Often ACM DL / CEUR-WS	Visibility and industry contacts
OARS (Offline Evaluation)	ACM RecSys	Offline evaluation, counterfactuals	Varies by edition	Evaluation methodology papers
HCREC (Human-Centered Rec)	ACM RecSys	User studies, interaction, transparency	Varies by edition	HCI-leaning recommender work
MLRec	SIAM SDM	ML/DM algorithms for recsys	Workshop proceedings	Algorithmic and data-mining depth
Domain-specific workshops	Various	News, music, e-commerce, health	Varies	Applied and domain papers

Treat this table as a starting framework, not a fixed ranking: workshop names, scope, and procedural policies change from year to year, so always confirm them against the current call for papers.

How to Evaluate a Workshop Before You Submit

1. Check the Proceedings Policy

Ask directly: Is the workshop archived? If so, where: ACM Digital Library, CEUR-WS, or Springer volume? An archival workshop paper may be cited and counts toward a publication record; a non-archival document is a discussion artifact. Neither is wrong, but the choice must be deliberate.

2. Read the Review Process

Some workshops use full peer review with three reviewers per article; others use light editing or acceptance of abstracts only. The list of program committees is a useful signal: a committee full of recognized names in your subfield usually means serious consideration. Look to see if the call mentions rebuttals, coaching, or a two-round process.

3. Match Scope to Your Contribution

An article on a new sequential recommendation architecture is part of a general RecSys or MLRec workshop. An article on how users perceive explanations corresponds to HCREC. An article on unbiased offline assessment corresponds to OARS. Inadequate scope is the most common reason why good articles get low reviews.

4. Weigh Logistics

Co-located workshops let you attend the main conference with one trip. Standalone workshops may be cheaper but offer less networking. For PhD students, the co-location and the main-conference discount often decide the question.

5. Consider the Audience You Need

If your goal is industry adoption, prioritize workshops with strong industry participation and a demonstration or industry track. If your goal is methodological feedback, favor smaller, theory-focused venues where the discussion is longer.

Related: — Deep, project-driven ML books and video courses — including the MEAP early-access program.

Unique Angles the Standard Lists Miss

Workshop papers can be extended into full papers. Many conferences explicitly allow a workshop paper to be extended into a conference submission, provided the extension is substantial (commonly cited as around 30% new content, though policies differ). This makes a workshop a low-risk way to get early feedback before a full submission.

Non-archival does not mean invisible. Non-archival workshops often still post accepted papers on the workshop site or on arXiv. The citation path is just different, and you should confirm with organizers how they expect the work to be referenced.

The “and” problem in scoping workshops. Workshops that attempt to cover too much - “user recommendation, modeling, evaluation and deployment” - tend to attract diffuse submissions and weaker reviews. Smaller workshops, such as those focused on a single evaluation question, often produce more focused discussions and better feedback.

Where we would start: — One-off, low-cost ML and recommender-systems courses you own forever.

Committee service is a signal. Serving on a program committee for a workshop you respect is one of the fastest ways to learn what makes a strong submission, and it builds the relationships that lead to collaboration. PhD students in particular should treat PC service as part of their training, not an obligation.

Reproducibility expectations are rising. Several recommender-systems workshops now encourage or require code and data release, and some run reproducibility tracks. A submission with a clean, runnable repository tends to fare better in review and gets cited more after publication.

A Practical Decision Framework

Start by naming your primary goal: visibility, feedback, or record. Then filter:

Visibility → RecSys workshops co-located with industry tracks and ACM DL procedures.
Feedback → smaller and specialized workshops (OARS, HCREC, MLRec) with long discussion periods.
Record → any workshop with confirmed archival proceedings and a documented review process.

Then check three practical constraints: deadline feasibility (can you produce a solid 4-8 page paper on time?), travel budget, and whether the workshop dates conflict with the sessions you need at the main conference. A workshop that you can fully commit to beats a prestigious workshop that you attend for two hours.

Sources & Further Reading

Recommender system — Wikipedia: A recommender system, also called a recommendation engine or content discovery platform is a type of information filtering system that aims to suggest items most…

Frequently Asked Questions

What is the best recommender systems workshop to submit to?

There is no better workshop: the right choice depends on your objective. For industry visibility and a citable record, ACM RecSys workshops collocated with archival proceedings are strong. For methodological feedback, specialized places like the OARS (offline evaluation) or the HCREC (human-centered recommendation) and the SIAM MLRec workshop offer more in-depth discussions. Match reach first, then prestige.

Are workshop papers worth it for a PhD student?

Workshop papers are often worthwhile for PhD students because they provide rapid, focused feedback and a low-risk first submission. They also build a publication record and introduce you to the community before attempting a full conference paper. The main caveat is that non-archived workshop materials carry less weight than archival materials, so check the proceedings policy.

How long should a recommender systems workshop paper be?

Most recommender-systems workshop papers run 4–8 pages in the conference’s standard format, though some accept extended abstracts of 2 pages and others allow up to 10. The exact limit is set by each workshop’s call for papers, so always confirm the current year’s requirements rather than relying on a previous edition.

Do workshop papers get published in the ACM Digital Library?

Some do and some do not. Workshops with archival proceedings may publish through the ACM Digital Library, CEUR-WS, or a Springer volume, while others are explicitly non-archival. The call for papers or workshop website states the policy, and it is the first thing to verify if a citable record matters to you.

Can I submit a paper to a workshop and a conference at the same time?

Most workshops prohibit the simultaneous submission of substantially identical work to an archival venue, and many treat a workshop paper as a precursor to a later, longer conference paper. Policies vary, so read the double submission clause in the call for papers and, if it is unclear, email the organizers before submitting.

What is the difference between MLRec and RecSys workshops?

MLRec is a data mining and machine learning workshop typically co-located with the SIAM International Conference on Data Mining, geared toward algorithms, scalability, and evaluation methodology. The RecSys workshops are co-hosted with the ACM Recommender Systems conference and focus more on applications, industrial deployment and human-centric topics. Both are legitimate; suitability depends on whether your contribution is methodological or applied.

Authoritative Sources to Consult

The ACM Recommender Systems conference workshop index: https://recsys.acm.org/workshops/
The RecSys 2024 workshops listing: https://recsys.acm.org/recsys24/workshops/
The OARS workshop on offline evaluation: https://oars-workshop.github.io/
The HCREC workshop on human-centered recommendation: https://hcrec.github.io/
A maintained community list of recommender-systems workshops: https://recommender-systems.com/publish/workshops/

For the underlying methodology, the Recommender Systems Handbook (Springer) and the ACM RecSys procedures remain the standard references.

P.S. A few readers have asked which marketplace courses we actually reach for — it's Udemy; if you want the current details.

Frequently asked questions