Sufficient consensus
Learning to rank trains a model to order documents by relevance, using features from the query, the document and past interactions. Pointwise, pairwise and listwise approaches are the standard families.
It replaced hand-tuned formulas in every large search system.
What the result pages leave out
These belong to the subject. People ask about them. They are missing from the agreed coverage.
- The feedback loop. Documents that were never shown generate no clicks, so they stay unlearnable. Position creates the data that confirms the position.
- What counts as a good document. Relevance labels come from raters working to a definition. The definition is a decision, and it is rarely published next to the result.
- New documents. A page with no interaction history starts with the weakest possible evidence, regardless of quality.
- Why a page lost. The model produces an order. It does not produce a reason that a publisher could act on.
- Optimization against the model. Once features are known, they are targeted. The model then measures the targeting rather than the quality.
What people actually want to know
- If my page has never been shown, how is it supposed to earn the data that would get it shown?
- Who wrote the definition of relevance that decides my visibility?
- What did the pages above me do that I can see, and what did they do that I cannot?
- How long does a new page need before the system has an opinion about it?
A learned ranking is a record of what was already visible. Publishing something that was never visible is the one move it cannot anticipate.
More in information retrieval
Snippet generationIt is written by a machine, from your text, for a query you did not know about.Query expansionYou are answering a question that was quietly edited before it reached you.Relevance feedbackAbsence produces no signal.Query performance predictionA page of ten results looks the same whether the answer exists or not.Passage retrievalYour page is being read in fragments by a system that never sees the whole.Mean average precisionA score of 0.87 tells you how well the ranking matched a list somebody wrote down.Vector space modelIf the web never described your case, the vector for it is somebody else's.Inverted indexThe web is the part of knowledge that somebody bothered to publish and a crawler managed to reach.