: Vocabulaire du TAL: search space

NLP resources and evaluation > data set > search space

search space

The set of all possible solutions that a search algorithm can explore to find the optimal solution for a given problem. In tasks like machine translation or text generation, the search space encompasses all the possible combinations of words or phrases that could form valid translations or generated text.

However again as with many other practical optimization problems the search space turns out to be intractably large. (Hatzivassiloglou & McKeown, 1993)
Since our solution space is the powerset of the considered features it can be seen as a lattice i.e. a graph where each solution is represented by a node. (Lion-Bouton, Savary & Antoine, 2023)
The search space is explored by beam search and model parameters are estimated with the averaged structured perceptron (Collins 2002) from training data consisting of question-answer pairs using answer F 1 -score as the objective. (Reddy, Täckström, Petrov, Steedman & Lapata, 2017)
These strategies navigate the solution space by selecting sequences of tokens that maximize the overall likelihood. (Massarelli, Petroni, Piktus, Ott, Rocktäschel, Plachouras, Silvestri & Riedel, 2020)
This knowledge sufficiently reduces the search space so that probabilistic parameters can be relatively easily estimated. (Nederhof & Rahman, 2015)

http://data.loterre.fr/ark:/67375/8LP-XMC0FD3V-H

RDF/XML TURTLE JSON-LD Dernière modification le 01/07/2024

Loterre