Jev canonical label mapping maps an old product label to a concept with a stable ID. Try closet organizers, pet grooming brushes or backpacking tents in the window. The supplied dictionary defines what each concept means. The returned label selects its exact local JSON record, including the demo taxonomy version.
An API-collected X post described tagging, deduplication and taxonomy normalization in an advertising workflow. This page implements the normalization decision independently. Its six labels and three concepts are original fixtures. The post's advertising volume, spending and expected product tiers are outside these tests.
A useful mapping preserves meaning across vocabulary changes. Two labels can sound different and still refer to the same product purpose. Conversely, a familiar word can name different things: a carrier for a pet belongs to Pet care in this dictionary, while an unexplained carrier needs clarification. The description travels with the label so the model has the evidence needed to distinguish those cases.