This paper provides an analysis of state-of-the-art activation functions with respect to supervised classification deep neural network. These comprise Rectified Linear Units (ReLU), Exponential Unit (ELU), Scaled (SELU), Gaussian Error (GELU), and the Inverse Square Root (ISRLU). To evaluate, experiments over two learning network architectures integrating these are conducted. The first model, based on Multilayer Perceptron (MLP), is evaluated MNIST dataset perform functions. Meanwhile,...

Determining the job is suitable for a student or person looking work based on their descriptions such as knowledge and skills that are difficult, well how employers must find ways to choose candidates match they require. In this paper, we focus studying prediction using different deep neural network models including TextCNN, Bi-GRU-LSTM-CNN, Bi-GRU-CNN with various pre-trained word embeddings IT dataset. addition, proposed simple effective ensemble model combining models. Our experimental...

In recent years, Hate Speech Detection has become one of the interesting fields in natural language processing or computational linguistics. this paper, we present description our system to solve problem at VLSP shared task 2019: on Social Networks with corpus which contains 20,345 human-labeled comments/posts for training and 5,086 public-testing. We implement a deep learning method based Bi-GRU-LSTM-CNN classifier into task. Our result is 70.576% F1-score, ranking 5th performance public-test set.

Although Vietnamese is the 17 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">th</sup> most popular native-speaker language in world, there are not many research studies on machine reading comprehension (MRC), task of understanding a text and answering questions about it. One reasons because lack high-quality benchmark datasets for this task. In work, we construct dataset which consists 2,783 pairs multiple-choice answers based 417 texts...

Machine reading comprehension is a natural language understanding task where the computing system required to read text and then find answer specific question posed by human. Large-scale high-quality corpora are necessary for evaluating machine models. Furthermore, (MRC) health sector has potential practical applications; nevertheless, MRC research in this domain currently scarce. This article presents UIT-ViNewsQA, new corpus Vietnamese evaluate models healthcare textual domain. The...

Speech emotion recognition is a crucial problem manifesting in multitude of applications such as human computer interaction and education. Although several advancements have been made the recent years, especially with advent Deep Neural Networks (DNN), most studies literature fail to consider semantic information speech signal. In this paper, we propose novel framework that can capture both paralinguistic particular, our comprised feature extractor, captures information, information. Both...

In this paper, we describe our system which participates in the shared task of Hate Speech Detection on Social Networks VLSP 2019 evaluation campaign. We are provided with pre-labeled dataset and an unlabeled for social media comments or posts. Our mission is to pre-process build machine learning models classify comments/posts. report, use Bidirectional Long Short-Term Memory model that can predict labels text according Clean, Offensive, Hate. With system, achieve comparative results 71.43%...

As biological gender is one of the aspects presenting individual human, much work has been done on classification based people names. The proposals for English and Chinese languages are tremendous; still, there have few works Vietnamese so far. We propose a new dataset prediction This comprises over 26,000 full names annotated with genders. available our website research purposes. In addition, this paper describes six machine learning algorithms (Support Vector Machine, Multinomial Naive...

Machine Reading Comprehension has attracted significant interest in research on natural language understanding, and large-scale datasets neural network-based methods have been developed for this task. However, most developments of resources machine reading comprehension investigated using two resource-rich languages, English Chinese. This article proposes a system called ViReader open-domain Vietnamese by Wikipedia as the textual knowledge source, where answer to any particular question is...

Large-scale and high-quality corpora are necessary for evaluating machine reading comprehension models on a low-resource language like Vietnamese. Besides, (MRC) the health domain offers great potential practical applications; however, there is still very little MRC research in this domain. This paper presents ViNewsQA as new corpus Vietnamese to evaluate healthcare models. The comprises 22,057 human-generated question-answer pairs. Crowd-workers create questions their answers based...

Machine reading comprehension (MRC) is a challenging task in natural language processing that makes computers understanding texts and answer questions based on those texts. There are many techniques for solving this problems, word representation very important technique impact most to the accuracy of machine problem popular languages like English Chinese. However, few studies MRC have been conducted low-resource such as Vietnamese. In paper, we conduct several experiments neural...

This paper provides an analysis of state-of-the-art activation functions with respect to supervised classification deep neural network. These comprise Rectified Linear Units (ReLU), Exponential Unit (ELU), Scaled (SELU), Gaussian Error (GELU), and the Inverse Square Root (ISRLU). To evaluate, experiments over two learning network architectures integrating these are conducted. The first model, basing on Multilayer Perceptron (MLP), is evaluated MNIST dataset perform functions.Meanwhile,...

Recent researches have demonstrated that BERT shows potential in a wide range of natural language processing tasks. It is adopted as an encoder for many state-of-the-art automatic summarizing systems, which achieve excellent performance. However, so far, there not much work done Vietnamese. In this paper, we showcase how can be implemented extractive text summarization Vietnamese on multi-document. We introduce novel comparison between different multilingual and monolingual models. The...