tvecs.evaluation package

Submodules

tvecs.evaluation.evaluation module

Module to Evaluate T-Vecs model against Human Semantic Similarity Score.

tvecs.evaluation.evaluation.extract_correlation_coefficient(score_data_path, vsm)

Extract Human Score, Word1, Word2. Compute T-Vecs Score.

API Documentation

param score_data_path

File generated by preprocessor/yandex

param vsm

Vector spaces mapped using 2 models.

type score_data_path

String

type vsm

tvecs.vector_space_mapper.vector_space_mapper

return

Returns (Correlation coefficient, P-Value)

rtype

Tuple(Float, Float)

tvecs.evaluation.evaluation.get_correlation_coefficient(human_score, calculated_score)

Measure correlation using Pearson’s Coefficient.

• The correlation is between the T-Vecs Model and

• Human Semantic Similarity Score.

API Documentation:

param human_score

List of human scores.

param calculated_score

List of calculated scores.

type human_score

List

type calculated_score

List

return

(Correlation Coefficient, P-Value)

rtype

Tuple(Float, Float)

Note

• correlation_coefficient - Measure of degree of relatedness between two variables

• p-value - The null hypothesis is that the two variables are uncorrelated. The p-value is a number between zero and one that represents the probability that your data would have arisen if the null hypothesis were true.

See also

• scipy.stats

tvecs.evaluation.preprocess_dataset module

Preprocess Evaluation Dataset by translating 1 column.

tvecs.evaluation.preprocess_dataset.preprocess_dataset(dataset_path, delimiter='', encoding='utf-8')

Preprocess Evaluation dataset by preprocessing 1 column.

Module contents