Retrieval models are often composed of two sub-models:

General Discussion
, ,
#1

I am reading tensorflow docs and I see that Retrieval models are often composed of two sub-models:

  1. A query model computing the query representation (normally a fixed-dimensionality embedding vector) using query features.
  2. A candidate model computing the candidate representation (an equally-sized vector) using the candidate features

I can’t understand what mean query model and candidate model as well as query representation and candidate representation. The outputs of the two models are then multiplied together to give a query-candidate affinity score, with higher scores expressing a better match between the candidate and the query.

Anyone may help me?