GeneralizedVector

class matchup.models.algorithms.generalized_vector.GeneralizedVector

Bases: matchup.models.model.IterModel

Class that implements the ‘run’ method of Generalized Vector IR model.

Methods Summary

generalized_calculate(doc, query) Calculate the similarity based on cosine of two vectors: doc vector and query vector.
generalized_doc_repr(term_repr, …) Generate vectors for all documents.
generalized_query_repr(query_weight, float], …) Generate query vector.
run(query, vocabulary) Run generalized vector model.
term_repr(minterms) Generate vectors for all terms based in minterms.

Methods Documentation

classmethod generalized_calculate(doc: List[float], query: List[float]) → float
Calculate the similarity based on cosine of two vectors: doc vector and query vector.
Parameters:
  • doc – doc vector
  • query – query vector
Returns:

score of doc
generalized_doc_repr(term_repr: DefaultDict[str, List[float]], base_len: int) → DefaultDict[str, List[float]]
Generate vectors for all documents.
Parameters:
  • term_repr – vector of all keys
  • base_len – len of these vectors
Returns:

All document representations
classmethod generalized_query_repr(query_weight: DefaultDict[str, float], term_repr: DefaultDict[str, List[float]], base_len: int) → List[float]
Generate query vector.
Parameters:
  • query_weight – query weighting based in keys.
  • term_repr – term vectors
  • base_len – base len of term vectors
Returns:
run(query: List[matchup.presentation.text.Term], vocabulary: matchup.structure.vocabulary.Vocabulary) → List[matchup.structure.solution.Result]
Run generalized vector model.
Parameters:
  • query – List of terms.
  • vocabulary – Vocabulary with a collection.
Returns:

Query results
classmethod term_repr(minterms: List[matchup.models.algorithms.generalized_vector.Minterm]) → Tuple[DefaultDict[str, List[float]], int]
Generate vectors for all terms based in minterms.
Parameters:minterms –
Returns: