How Are Computer Science Conferences Ranked?

Computer science conferences are often more important than journals for announcing new results, but they are not ranked by one universally accepted system. A conference that is prestigious in computer architecture may be little known in human–computer interaction, while a workshop valued in HCI may not matter in theoretical computer science.

The useful question is therefore not simply “What is the global ranking?” but “How is this venue regarded by researchers in my subfield, and what does that reputation represent?”

There is no universal ranking

Conference quality is multidimensional. A ranking may combine several different ideas:

• Reputation: whether respected researchers consider the venue influential.
• Selectivity: how many submissions are accepted.
• Scientific quality: the strength and originality of the accepted work.
• Visibility: whether the relevant research community attends and reads it.
• Review standards: the rigor, transparency, and expertise of peer review.
• Practical value: opportunities for discussion, collaboration, and feedback.

These properties do not always coincide. A small specialist conference can have excellent papers and an influential audience without being highly selective. Conversely, a large conference can have a low acceptance rate partly because it receives many submissions, not necessarily because every accepted paper is better than work elsewhere.

Common ranking systems

Several lists are used in particular countries or disciplines. They should be treated as indicators rather than official scientific truth.

CORE rankings

The Computing Research and Education Association of Australasia (CORE) historically classified computing conferences into categories such as A*, A, B, and C. These lists are widely encountered in academic evaluations, but their coverage and currency vary by field and by edition. Because any CORE ranking is tied to a particular edition, check its date and current status before relying on it. A category is not a substitute for checking how researchers in your area currently view the venue.

CCF rankings

The China Computer Federation (CCF) maintains a classification of selected conferences and journals in different computing areas. It is influential in some institutional and national evaluation systems, especially in China. Its categories are not a universal standard for hiring, promotion, or admissions elsewhere.

Informal community lists

Departments, research groups, and researchers often maintain informal lists such as “top-tier,” “strong,” or “regional” venues. These can be more useful than a global list when they are recent and specific to a subfield. Their weakness is that they may reflect one institution’s incentives or an individual’s preferences.

Bibliometric indicators

Citation counts, h-indexes, and proceedings metrics can show visibility, but they are affected by field size, publication age, self-citation, and indexing coverage. They should not be interpreted as direct measurements of the quality of an individual paper or conference.

How to evaluate a conference yourself

Use several checks rather than relying on one label.

1. Identify the relevant subfield

Start with the exact area: for example, machine learning, programming languages, databases, computer networks, security, graphics, or theory. Search for venues where researchers publishing work similar to yours regularly publish. A broad computer-science ranking can hide major differences between communities.

2. Inspect recent programs

Look at the last two or three programs. Are the papers addressing important questions? Do recognizable researchers in the field publish there? Are the papers cited, discussed, or extended by later work? A program gives more current information than an old ranking.

3. Examine the review model

Check whether the conference uses a single-round or multi-round process, double-blind review, rebuttals, shepherding, artifact evaluation, or other quality controls. None of these mechanisms guarantees good decisions, but they help you understand what acceptance means.

Also check what is actually published. Some venues publish full archival papers; others publish short papers, abstracts, or presentations without archival proceedings. Neither model is automatically better, but they serve different purposes and may be treated differently by employers and universities.

4. Interpret acceptance rates carefully

An acceptance rate is commonly calculated as

$$\text{acceptance rate} = \frac{\text{number of accepted papers}}{\text{number of submissions counted by the venue}}.$$

For example, 240 accepted papers out of 1,200 counted submissions gives an acceptance rate of $240/1200=20\%$. This indicates selectivity, but not by itself scientific quality. Depending on the venue’s counting rules, submission numbers may include incomplete, out-of-scope, or low-effort papers, and conferences differ in paper length, tracks, and counting rules.

5. Ask active researchers

A faculty member, postdoctoral researcher, or senior PhD student who publishes recently in your area can often answer questions that rankings cannot: whether the venue is taken seriously, whether the review process is constructive, and whether the audience is appropriate for your work.

Ask about a specific comparison—for example, “How is Conference A viewed relative to Conference B for systems papers?”—rather than asking for a single universal ordering.

A worked comparison

Suppose you are choosing between three venues for a database paper:

Signal	Venue A	Venue B	Venue C
Recent acceptance rate	18%	32%	12%
Relevant papers in recent programs	Many	Some	Few
Target community attends	Yes	Yes	Unclear
Archival full papers	Yes	Yes	No
Review feedback	Detailed	Variable	Unknown

Venue C should not automatically win because its acceptance rate is lowest. If the relevant database community rarely attends and the venue publishes no archival full papers, it may be a poor fit. Venue A may be the strongest choice if its audience, papers, and review process match the project. Venue B may be appropriate if it offers a broader audience or a better submission timeline.

The best venue depends on your goal. For a major research contribution, visibility among the right specialists may matter most. For early feedback, a workshop or smaller conference may be more useful. For a permanent archival record, publication format and indexing deserve particular attention.

Common mistakes

• Treating a ranking as timeless. Leadership, scope, review quality, and reputation can change.
• Comparing different fields directly. “Top” in theory, AI, and HCI does not mean the same thing.
• Using acceptance rate as a quality score. Selectivity measures competition, not the complete value of the venue.
• Confusing a conference with a paper. A strong conference can publish a weak paper, and an excellent result can appear in a less prestigious venue.
• Ignoring publication rules. Some departments count archival conference papers differently from journals or non-archival presentations.
• Submitting only because of a label. A good match between the paper and the conference’s audience improves both review and impact.

For a CV or an academic application, describe the venue accurately and provide context when appropriate—such as the subfield, whether the paper was peer reviewed, and whether the proceedings are archival. Do not claim that a conference is “number one” unless a specific, relevant organization defines that ranking and the claim is still current.