Large-scale Translation to Enable Response Selection in Low Resource Languages: A COVID-19 Chatbot Experiment

Aguiar, Lucas Almeida; Cruz, Lívia Almada; Silva, Ticiana L. Coelho da; Carmo, Rafael Augusto Ferreira do; Paixao, Matheus Henrique Esteves

doi:10.5753/sbbd.2022.224329

Cited by 1 publication

(1 citation statement)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The work [52] applies data augmentation to increase training data to response selection in chatbots based on multi-turn recovering. They apply the automatic translation of a massive dataset for multi-turn chatbots from English to Brazilian Portuguese, train a deep neural network with the translated dataset, and tune the neural network using a COVID-19-related dialogues dataset.…”

Section: Related Workmentioning

confidence: 99%

Learning and Semi-automatic Intention Labeling for Classification Models: ACOVID-19 Study for Chatbots

Júnior

Oliveira

Cruz

et al. 2022

Preprint

View full text Add to dashboard Cite

It is increasingly common to use chatbots as an interface to services. Making this experience more humanized requires the chatbot to understand natural language and express itself using it. One of the main components of a chatbot is Natural Language Understanding (NLU) model, which is responsible for interpreting the text and extracting the intent and entities present in that text. It's possible to focus only on one of these tasks of NLU, such as Intent classification. To train an NLU intent classification model, it's generally necessary to use a considerable amount of annotated data, where each sentence of the dataset receives a label indicating an intent.Performing manually labeling data is an arduous and impracticable process depending on the data volume. Thus, an unsupervised machine learning technique, such as data clustering, could be applied to find patterns in the data and label them. For this task, it is essential to have an effective vector embedding representation of texts that depicts the semantic information and helps the machine understand the context, intent, and other nuances of the entire text. In this paper, we perform an extensive evaluation of different text embedding models for clustering and labeling, we also apply some operations to improve the quality of the dataset, such as removing sentences establishing different strategies for distance thresholds in terms of (cosine similarity) for the clusters' centroids. Then we trained some Intent Classification Models with two different architectures, one built with the Rasa framework and the other with a Neural Network (NN) using the text of attendance from the Coronavirus Platform Service of Ceará, Brazil. We also manually annotated a dataset to be used as validation data. We found that semi-automatic labeling through clustering and visual inspection added some biases to the intent classification models. However, we still achieved competitive results in terms of accuracy for the trained models.

show abstract

Section: Related Workmentioning

confidence: 99%