<code>afe()</code> computes Average Feature Entropy (AFE), which measures randomness of
occurrences of features in labelled documents (Watanabe &amp; Zhou, 2020). In
creating seed dictionaries, AFE can be used to avoid adding seed words that would
decrease classification accuracy.

Extract features and classify documents with noisy labels given by document-meta data or keyword matching Watanabe & Zhou (2020) <doi:10.1177/0894439320907027>.

Kohei Watanabe

wordmap

Feature Extraction and Document Classification with Noisy Labels

afe function

<dl><dt>x</dt>
<dd>a dfm for features.</dd>
<dt>y</dt>
<dd>a dfm for labels.</dd>
<dt>smooth</dt>
<dd>a numeric value for smoothing to include all the features.</dd></dl>

Arguments

Compute Average Feature Entropy (AFE) — afe

<dl>

<dt>x</dt>
<dd>a dfm for features.</dd>


<dt>y</dt>
<dd>a dfm for labels.</dd>


<dt>smooth</dt>
<dd>a numeric value for smoothing to include all the features.</dd>

</dl>

afe: Compute Average Feature Entropy (AFE)

Description

Usage

Value

Arguments

References