VLDB 2026 Research / matching

Assignment

Choose where the matching scores come from, set the chairs' options, then solve.
Export CSV
⇪1 · Import existing matching scores

Bring the scores you already have (e.g. from bids, TPMS, or a previous cycle) and let TaxoReview do the assignment only. CSV or JSON rows with sub_id, rev_id, score; higher is better. sample

⌥2 · Match and assign from scratch

TaxoReview embeds every text with SPECTER2, builds the local weighted taxonomy over submissions and reviewer evidence, computes taxonomy affinities, and then solves the assignment with the options below.

Assignment options — shared by both routes; solve re-uses the current score source

Reviewers per submission
Every paper gets between min and max reviewers.
Submissions per reviewer
Per-reviewer caps set on the Reviewers page override max.
Balance threshold
Every assigned pair must score at least this; papers with too few eligible reviewers are reported.
Diversity & independence 278/304 submissions and 0/400 reviewers carry a country; unknown countries are never blocked. From the reviewers' publication lists, plus the expert database's authorship graph for the 400 reviewers linked to it.
Import scores or run the matching first.
⇄
Import matching scores or run the matching from scratch, then solve.