Human annotated model

The human-annotated model is a composition of two functions, i.e., \(g \circ m_b\) where \(m_b\) is created using the procedure described in Arabic, English, and Spanish for Arabic, English and Spanish, respectively. The classifier \(g\) is trained with human-annotated datasets obtained from Arabic Sentiment Analysis and Cross-lingual Sentiment Resources for Arabic, and for English and Spanish it is used the data from [Human-Annotated].

The model has the following signature \(g \circ m_b: \text{text} \rightarrow \mathbb R^3\), where the first component corresponds to the negative class, the second to the neutral, and the third to the positive.

This model can be used as follow, firstly let us read a dataset to train EvoMSA.

>>> from EvoMSA import base
>>> from microtc.utils import tweet_iterator
>>> import os
>>> tweets = os.path.join(os.path.dirname(base.__file__), 'tests', 'tweets.json')
>>> D = list(tweet_iterator(tweets))
>>> X = [x['text'] for x in D]
>>> y = [x['klass'] for x in D]

Once the dataset is load, EvoMSA using HA Space in Spanish is trained as follows:

>>> from EvoMSA.base import EvoMSA
>>> evo = EvoMSA(HA=True, lang='es').fit(X, y)
>>> evo.predict(['buenos dias'])