Aspect-based sentiment analysis has been studied in both research and industrial communities over recent years. For the low-resource languages, the standard benchmark corpora play an important role in the development of methods. In this article, we introduce two benchmark corpora with the largest sizes at sentence-level for two tasks: Aspect Category Detection and Aspect Polarity Classification in Vietnamese. Our corpora are annotated with high inter-annotator agreements for the restaurant and hotel domains. The release of our corpora would push forward the low-resource language processing community. In addition, we deploy and compare the effectiveness of supervised learning methods with a single and multi-task approach based on deep learning architectures. Experimental results on our corpora show that the multi-task approach based on BERT architecture outperforms the neural network architectures and the single approach. Our corpora and source code are published on this footnoted site. 1

Language Identification (LID) is a challenging task, especially when the input texts are short and noisy such as microblog posts on social media or chat logs on gaming forums. The task has been tackled by either designing a feature set for a traditional classifier (e.g. Naive Bayes) or applying a deep neural network classifier (e.g. Bi-directional GRU, Encoder-Decoder). These methods are usually trained and tested on a private corpus, then used as off-the-shelf packages by other researchers on their own datasets, and consequently the various results published are not directly comparable. In this paper, we first create a new massive labeled dataset based on one year of Twitter data. We use this dataset to test several existing LID systems, in order to obtain a set of coherent benchmarks, and we make our dataset publicly available so that others can add to this set of benchmarks. Finally, we propose a shallow but efficient neural LID system, which is a ngram-regional convolution neural network enhanced with an attention mechanism. Experimental results show that our architecture is able to predict tens of thousands of samples per second and surpasses all state-of-the-art systems in accuracy and F1 score, including outperforming the popular langid system by 5%.

Targeted sentiment analysis classifies the sentiment polarity towards each target entity mention in given text documents. Seminal methods extract manual discrete features from automatic syntactic parse trees in order to capture semantic information of the enclosing sentence with respect to a target entity mention. Recently, it has been shown that competitive accuracies can be achieved without using syntactic parsers, which can be highly inaccurate on noisy text such as tweets. This is achieved by applying distributed word representations and rich neural pooling functions over a simple and intuitive segmentation of tweets according to target entity mentions. In this paper, we extend this idea by proposing a sentence-level neural model to address the limitation of pooling functions, which do not explicitly model tweet-level semantics. First, a bi-directional gated neural network is used to connect the words in a tweet so that pooling functions can be applied over the hidden layer instead of words for better representing the target and its contexts. Second, a three-way gated neural network structure is used to model the interaction between the target mention and its surrounding contexts. Experiments show that our proposed model gives significantly higher accuracies compared to the current best method for targeted sentiment analysis.

Sentiment lexicons have been leveraged as a useful source of features for sentiment analysis models, leading to the state-of-the-art accuracies.On the other hand, most existing methods use sentiment lexicons without considering context, typically taking the count, sum of strength, or maximum sentiment scores over the whole input.We propose a context-sensitive lexicon-based method based on a simple weighted-sum model, using a recurrent neural network to learn the sentiments strength, intensification and negation of lexicon sentiments in composing the sentiment value of sentences.Results show that our model can not only learn such operation details, but also give significant improvements over state-of-the-art recurrent neural network baselines without lexical features, achieving the best results on a Twitter benchmark.

Open domain targeted sentiment is the joint information extraction task that finds target mentions together with the sentiment towards each mention from a text corpus.The task is typically modeled as a sequence labeling problem, and solved using state-of-the-art labelers such as CRF.We empirically study the effect of word embeddings and automatic feature combinations on the task by extending a CRF baseline using neural networks, which have demonstrated large potentials for sentiment analysis.Results show that the neural model can give better results by significantly increasing the recall.In addition, we propose a novel integration of neural and discrete features, which combines their relative advantages, leading to significantly higher results compared to both baselines.