Clustering Vietnamese Conversations From Facebook Page To Build Training Dataset For Chatbot

TL;DR

This paper presents a method to extract and cluster Vietnamese Facebook conversations using PhoBERT embeddings to efficiently generate training data for chatbots, demonstrating improved clustering performance and dataset creation.

Contribution

The study introduces a novel approach combining Facebook data extraction, PhoBERT-based feature extraction, and clustering algorithms with parameter optimization for Vietnamese chatbot training data.

Findings

PhoBERT outperforms other models in feature extraction.

Clustering algorithms achieve high V-measure and Silhouette scores.

The method significantly reduces time and effort in dataset creation.

Abstract

The biggest challenge of building chatbots is training data. The required data must be realistic and large enough to train chatbots. We create a tool to get actual training data from Facebook messenger of a Facebook page. After text preprocessing steps, the newly obtained dataset generates FVnC and Sample dataset. We use the Retraining of BERT for Vietnamese (PhoBERT) to extract features of our text data. K-Means and DBSCAN clustering algorithms are used for clustering tasks based on output embeddings from PhoBERT$_{base}$. We apply V-measure score and Silhouette score to evaluate the performance of clustering algorithms. We also demonstrate the efficiency of PhoBERT compared to other models in feature extraction on the Sample dataset and wiki dataset. A GridSearch algorithm that combines both clustering evaluations is also proposed to find optimal parameters. Thanks to clustering such a number of conversations, we save a lot of time and effort to build data and storylines for training chatbot.