Text-to-hashtag Generation using Seq2seq Learning

TL;DR

This paper investigates the use of BiLSTM and BERT models to predict hashtags for Ecommerce reviews in Brazilian Portuguese, showing promising results through crowdsourced evaluation despite poor traditional metric scores.

Contribution

It introduces a seq2seq approach using BiLSTM and BERT for hashtag prediction in Ecommerce, highlighting the effectiveness of neural networks validated by crowdsourced assessments.

Findings

Crowdsourced scores indicate high quality of predicted hashtags.

Traditional metrics (NIST, BLEU, METEOR) show poor results.

Neural network-generated hashtags are promising for Ecommerce applications.

Abstract

In this paper, we studied whether models based on BiLSTM and BERT can predict hashtags in Brazilian Portuguese for Ecommerce websites. Hashtags have a sizable financial impact on Ecommerce. We processed a corpus of Ecommerce reviews as inputs, and predicted hashtags as outputs. We evaluated the results using four quantitative metrics: NIST, BLEU, METEOR and a crowdsourced score. A word cloud was used as a qualitative metric. While all computer-generated metrics (NIST, BLEU and METEOR) indicated bad results, the crowdsourced results produced amazing scores. We concluded that the texts predicted by the neural networks are very promising for use as hashtags for products on Ecommerce websites. The code for this work is available at https://github.com/augustocamargo/text-to-hashtag.