crowdkit.aggregation.texts.text_hrrasa.TextHRRASA | Source code

    encoder: Callable,
    n_iter: int = 100,
    tol: float = 1e-05,
    lambda_emb: float = 0.5,
    lambda_out: float = 0.5,
    alpha: float = 0.05,
    calculate_ranks: bool = False,
    output_similarity: Callable = glue_similarity

HRRASA on text embeddings.

Given a sentence encoder, encodes texts provided by workers and runs the HRRASA algorithm for embedding aggregation.

Parameters Description

Parameters Type Description
encoder Callable

A callable that takes a text and returns a NumPy array containing the corresponding embedding.

n_iter int

A number of HRRASA iterations.

lambda_emb float

A weight of reliability calculated on embeddigs.

lambda_out float

A weight of reliability calculated on outputs.

alpha float

Confidence level of chi-squared distribution quantiles in beta parameter formula.

calculate_ranks bool

If true, calculate additional attribute ranks_.


We suggest to use sentence encoders provided by Sentence Transformers.

from crowdkit.datasets import load_dataset
from crowdkit.aggregation import TextHRRASA
from sentence_transformers import SentenceTransformer
encoder = SentenceTransformer('all-mpnet-base-v2')
hrrasa = TextHRRASA(encoder=encoder.encode)
df, gt = load_dataset('crowdspeech-test-clean')
df['text'] = df['text'].apply(lambda s: s.lower())
result = hrrasa.fit_predict(df)

Methods Summary

Method Description
fit_predict Fit the model and return aggregated texts.
fit_predict_scores Fit the model and return scores.