Search

In this paper, we propose a novel Human-like Semantic Cognition Network (HSCN) for aspect-level sentiment classification, motivated by the principles of human beings’ reading cognitive process (pre-reading, active reading, post-reading). We first design a word-level interactive perception module to capture the correlation between context words and the given target words, which can be regarded as pre-reading. Second, to mimic the process of active reading, we propose a targetaware semantic distillation module to produce the targetspecific context representation for aspect-level sentiment prediction. Third, we further devise a semantic deviation metric module to measure the semantic deviation between the targetspecific context representation and the given target, which evaluates the degree we understand the target-specific context semantics. The measured semantic deviation is then used to fine-tune the above active reading process in a feedback regulation way. To verify the effectiveness of our approach, we conduct extensive experiments on three widely used datasets. The experiments demonstrate that HSCN achieves impressive results compared to other strong competitors.

In the last decades, many similarity measures are proposed, such as Jaccard coefficient, cosine similarity, BM25, language model, etc. Despite the effectiveness of the existing similarity measures, we observe that none of them can consistently outperform the others in most typical situations. Choosing which similarity predicate to use is usually treated as an empirical question by evaluating a particular task with a number of different similarity predicates, which is not computationally efficient and the obtained results are not portable. In this paper, we propose a novel approach to combine different similarity predicates together to form a committee so that we do not need to worry about choosing which of them to use. Empirically, we can obtain a better result than any individual similarity predicate, which is quite meaningful in practice. Specifically, our method models the problem of committee generation as a 0–1 integer programming problem based on the confidence of similarity predicates and the reliability of attributes. We demonstrate the effectiveness of our model by applying it on three datasets with controlled errors. Experimental results demonstrate that our similarity predicate committee is more robust and superior over existing individual similarity predicates.

With the transformation of education from the traditional classroom environment to online education and assessment, it is more and more important to accurately assess the difficulty of questions than ever. As teachers may not be able to follow the student’s performance and learning behavior closely, a well-defined method to measure the difficulty of questions to guide learning is necessary. In this paper, we explore the concept of question difficulty and provide our new Chinese DT-QDC dataset. This is currently the largest and only Chinese question dataset, and it also has enriched attributes and difficulty labels. Additional attributes such as keywords, chapter, and question type would allow models to understand questions more precisely. We proposed the MTMS-BERT and ORMS-BERT, which can improve the judgment of difficulty from different views. The proposed methods outperforms different baselines by 7.79% on F1-score and 15.92% on MAE, 28.26% on MSE on the new DT-QDC dataset, laying the foundation for the question difficulty comprehension task.

Recent years have seen the dependency tree as effective information for relation extraction. Two problems still exist in previous methods: (1) dependency tree relies on external tools and needs to be carefully integrated with a trade-off between pruning noisy words and keeping semantic integrity; (2) dependency-based methods still have to encode sequential context as a supplement, which needs extra time. To tackle the two problems, we propose a faster dependency-free model in this paper: regarding the sentence as a fully-connected graph, we customize the vanilla transformer architecture to remove the irrelevant information via filtering mechanism and further aggregate the sentence information through the enhanced query. Our model yields comparable results on the SemEval2010 Task8 dataset and better results on the TACRED dataset, without requiring external information from the dependency tree but with improved time efficiency.

Deep learning approaches for sentiment classification do not fully exploit sentiment linguistic knowledge. In this paper, we propose a Multi-sentiment-resource Enhanced Attention Network (MEAN) to alleviate the problem by integrating three kinds of sentiment linguistic knowledge (e.g., sentiment lexicon, negation words, intensity words) into the deep neural network via attention mechanisms. By using various types of sentiment resources, MEAN utilizes sentiment-relevant information from different representation sub-spaces, which makes it more effective to capture the overall semantics of the sentiment, negation and intensity words for sentiment prediction. The experimental results demonstrate that MEAN has robust superiority over strong competitors.

Analyzing public opinions towards products, services and social events is an important but challenging task. Despite the remarkable successes of deep neural networks in sentiment analysis, these approaches do not make full use of the prior sentiment knowledge (e.g., sentiment lexicon, negation words, intensity words). In this paper, we propose a Sentiment-Aware Attention Network (SAAN) to boost the performance of sentiment analysis, which adopts a three-step strategy to learn the sentiment-specific sentence representation. First, we employ a word-level mutual attention mechanism to model word-level correlation. Next, a phrase-level convolutional attention is designed to obtain phrase-level correlation. Finally, a sentence-level multi-head attention mechanism is proposed to capture various sentimental information from different subspaces. The experiments on Movie Review (MR) and Stanford Sentiment Treebank (SST) show that our model consistently outperform the previous methods for sentiment analysis.

Sentiment analysis of social media and comment data is an important issue in opinion monitoring. In this work, we propose a Linguistic-Aware Attention Network (LANN) to enhance the performance of convolution neural network (CNN). LANN adopts a two-stage strategy to model the sentiment-specific sentence representation. First, an interactive attention mechanism is designed to model word-level semantics. Second, to capture phrase-level linguistic structure, a dynamic semantic attention is adopted to select the crucial phrase chunks in the sentence. The experiments demonstrate that LANN has robust superiority over competitors and has reached the state-of-the-art performance.

In this study, we explore capsule networks with dynamic routing for text classification. We propose three strategies to stabilize the dynamic routing process to alleviate the disturbance of some noise capsules which may contain "background" information or have not been successfully trained. A series of experiments are conducted with capsule networks on six text classification benchmarks. Capsule networks achieve state of the art on 4 out of 6 datasets, which shows the effectiveness of capsule networks for text classification. We additionally show that capsule networks exhibit significant improvement when transfer single-label to multi-label text classification over strong baseline methods. To the best of our knowledge, this is the first work that capsule networks have been empirically investigated for text modeling., 12 pages