Sixing Wu


pdf bib
Section-Aware Commonsense Knowledge-Grounded Dialogue Generation with Pre-trained Language Model
Sixing Wu | Ying Li | Ping Xue | Dawei Zhang | Zhonghai Wu
Proceedings of the 29th International Conference on Computational Linguistics

In knowledge-grounded dialogue generation, pre-trained language models (PLMs) can be expected to deepen the fusing of dialogue context and knowledge because of their superior ability of semantic understanding. Unlike adopting the plain text knowledge, it is thorny to leverage the structural commonsense knowledge when using PLMs because most PLMs can only operate plain texts. Thus, linearizing commonsense knowledge facts into plan text is a compulsory trick. However, a dialogue is always aligned to a lot of retrieved fact candidates; as a result, the linearized text is always lengthy and then significantly increases the burden of using PLMs. To address this issue, we propose a novel two-stage framework SAKDP. In the first pre-screening stage, we use a ranking network PriorRanking to estimate the relevance of a retrieved knowledge fact. Thus, facts can be clustered into three sections of different priorities. As priority decreases, the relevance decreases, and the number of included facts increases. In the next dialogue generation stage, we use section-aware strategies to encode the linearized knowledge. The powerful but expensive PLM is only used for a few facts in the higher priority sections, reaching the performance-efficiency balance. Both the automatic and human evaluation demonstrate the superior performance of this work.

pdf bib
KSAM: Infusing Multi-Source Knowledge into Dialogue Generation via Knowledge Source Aware Multi-Head Decoding
Sixing Wu | Ying Li | Dawei Zhang | Zhonghai Wu
Findings of the Association for Computational Linguistics: ACL 2022

Knowledge-enhanced methods have bridged the gap between human beings and machines in generating dialogue responses. However, most previous works solely seek knowledge from a single source, and thus they often fail to obtain available knowledge because of the insufficient coverage of a single knowledge source. To this end, infusing knowledge from multiple sources becomes a trend. This paper proposes a novel approach Knowledge Source Aware Multi-Head Decoding, KSAM, to infuse multi-source knowledge into dialogue generation more efficiently. Rather than following the traditional single decoder paradigm, KSAM uses multiple independent source-aware decoder heads to alleviate three challenging problems in infusing multi-source knowledge, namely, the diversity among different knowledge sources, the indefinite knowledge alignment issue, and the insufficient flexibility/scalability in knowledge usage. Experiments on a Chinese multi-source knowledge-aligned dataset demonstrate the superior performance of KSAM against various competitive approaches.

pdf bib
MOBA-E2C: Generating MOBA Game Commentaries via Capturing Highlight Events from the Meta-Data
Dawei Zhang | Sixing Wu | Yao Guo | Xiangqun Chen
Findings of the Association for Computational Linguistics: EMNLP 2022

MOBA (Multiplayer Online Battle Arena) games such as Dota2 are currently one of the most popular e-sports gaming genres. Following professional commentaries is a great way to understand and enjoy a MOBA game. However, massive game competitions lack commentaries because of the shortage of professional human commentators. As an alternative, employing machine commentators that can work at any time and place is a feasible solution. Considering the challenges in modeling MOBA games, we propose a data-driven MOBA commentary generation framework, MOBA-E2C, allowing a model to generate commentaries based on the game meta-data. Subsequently, to alleviate the burden of collecting supervised data, we propose a MOBA-FuseGPT generator to generate MOBA game commentaries by fusing the power of a rule-based generator and a generative GPT generator. Finally, in the experiments, we take a popular MOBA game Dota2 as our case and construct a Chinese Dota2 commentary generation dataset Dota2-Commentary. Experimental results demonstrate the superior performance of our approach. To the best of our knowledge, this work is the first Dota2 machine commentator and Dota2-Commentary is the first dataset.


pdf bib
More is Better: Enhancing Open-Domain Dialogue Generation via Multi-Source Heterogeneous Knowledge
Sixing Wu | Ying Li | Minghui Wang | Dawei Zhang | Yang Zhou | Zhonghai Wu
Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing

Despite achieving remarkable performance, previous knowledge-enhanced works usually only use a single-source homogeneous knowledge base of limited knowledge coverage. Thus, they often degenerate into traditional methods because not all dialogues can be linked with knowledge entries. This paper proposes a novel dialogue generation model, MSKE-Dialog, to solve this issue with three unique advantages: (1) Rather than only one, MSKE-Dialog can simultaneously leverage multiple heterogeneous knowledge sources (it includes but is not limited to commonsense knowledge facts, text knowledge, infobox knowledge) to improve the knowledge coverage; (2) To avoid the topic conflict among the context and different knowledge sources, we propose a Multi-Reference Selection to better select context/knowledge; (3) We propose a Multi-Reference Generation to generate informative responses by referring to multiple generation references at the same time. Extensive evaluations on a Chinese dataset show the superior performance of this work against various state-of-the-art approaches. To our best knowledge, this work is the first to use the multi-source heterogeneous knowledge in the open-domain knowledge-enhanced dialogue generation.

pdf bib
Adversarial Attack against Cross-lingual Knowledge Graph Alignment
Zeru Zhang | Zijie Zhang | Yang Zhou | Lingfei Wu | Sixing Wu | Xiaoying Han | Dejing Dou | Tianshi Che | Da Yan
Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing

Recent literatures have shown that knowledge graph (KG) learning models are highly vulnerable to adversarial attacks. However, there is still a paucity of vulnerability analyses of cross-lingual entity alignment under adversarial attacks. This paper proposes an adversarial attack model with two novel attack techniques to perturb the KG structure and degrade the quality of deep cross-lingual entity alignment. First, an entity density maximization method is employed to hide the attacked entities in dense regions in two KGs, such that the derived perturbations are unnoticeable. Second, an attack signal amplification method is developed to reduce the gradient vanishing issues in the process of adversarial attacks for further improving the attack effectiveness.


pdf bib
Improving Knowledge-Aware Dialogue Response Generation by Using Human-Written Prototype Dialogues
Sixing Wu | Ying Li | Dawei Zhang | Zhonghai Wu
Findings of the Association for Computational Linguistics: EMNLP 2020

Incorporating commonsense knowledge can alleviate the issue of generating generic responses in open-domain generative dialogue systems. However, selecting knowledge facts for the dialogue context is still a challenge. The widely used approach Entity Name Matching always retrieves irrelevant facts from the view of local entity words. This paper proposes a novel knowledge selection approach, Prototype-KR, and a knowledge-aware generative model, Prototype-KRG. Given a query, our approach first retrieves a set of prototype dialogues that are relevant to the query. We find knowledge facts used in prototype dialogues usually are highly relevant to the current query; thus, Prototype-KR ranks such knowledge facts based on the semantic similarity and then selects the most appropriate facts. Subsequently, Prototype-KRG can generate an informative response using the selected knowledge facts. Experiments demonstrate that our approach has achieved notable improvements on the most metrics, compared to generative baselines. Meanwhile, compared to IR(Retrieval)-based baselines, responses generated by our approach are more relevant to the context and have comparable informativeness.

pdf bib
Diverse and Informative Dialogue Generation with Context-Specific Commonsense Knowledge Awareness
Sixing Wu | Ying Li | Dawei Zhang | Yang Zhou | Zhonghai Wu
Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics

Generative dialogue systems tend to produce generic responses, which often leads to boring conversations. For alleviating this issue, Recent studies proposed to retrieve and introduce knowledge facts from knowledge graphs. While this paradigm works to a certain extent, it usually retrieves knowledge facts only based on the entity word itself, without considering the specific dialogue context. Thus, the introduction of the context-irrelevant knowledge facts can impact the quality of generations. To this end, this paper proposes a novel commonsense knowledge-aware dialogue generation model, ConKADI. We design a Felicitous Fact mechanism to help the model focus on the knowledge facts that are highly relevant to the context; furthermore, two techniques, Context-Knowledge Fusion and Flexible Mode Fusion are proposed to facilitate the integration of the knowledge in the ConKADI. We collect and build a large-scale Chinese dataset aligned with the commonsense knowledge for dialogue generation. Extensive evaluations over both an open-released English dataset and our Chinese dataset demonstrate that our approach ConKADI outperforms the state-of-the-art approach CCM, in most experiments.


pdf bib
HL-EncDec: A Hybrid-Level Encoder-Decoder for Neural Response Generation
Sixing Wu | Dawei Zhang | Ying Li | Xing Xie | Zhonghai Wu
Proceedings of the 27th International Conference on Computational Linguistics

Recent years have witnessed a surge of interest on response generation for neural conversation systems. Most existing models are implemented by following the Encoder-Decoder framework and operate sentences of conversations at word-level. The word-level model is suffering from the Unknown Words Issue and the Preference Issue, which seriously impact the quality of generated responses, for example, generated responses may become irrelevant or too general (i.e. safe responses). To address these issues, this paper proposes a hybrid-level Encoder-Decoder model (HL-EncDec), which not only utilizes the word-level features but also character-level features. We conduct several experiments to evaluate HL-EncDec on a Chinese corpus, experimental results show our model significantly outperforms other non-word-level models in automatic metrics and human annotations and is able to generate more informative responses. We also conduct experiments with a small-scale English dataset to show the generalization ability.

pdf bib
THU_NGN at SemEval-2018 Task 3: Tweet Irony Detection with Densely connected LSTM and Multi-task Learning
Chuhan Wu | Fangzhao Wu | Sixing Wu | Junxin Liu | Zhigang Yuan | Yongfeng Huang
Proceedings of the 12th International Workshop on Semantic Evaluation

Detecting irony is an important task to mine fine-grained information from social web messages. Therefore, the Semeval-2018 task 3 is aimed to detect the ironic tweets (subtask A) and their ironic types (subtask B). In order to address this task, we propose a system based on a densely connected LSTM network with multi-task learning strategy. In our dense LSTM model, each layer will take all outputs from previous layers as input. The last LSTM layer will output the hidden representations of texts, and they will be used in three classification task. In addition, we incorporate several types of features to improve the model performance. Our model achieved an F-score of 70.54 (ranked 2/43) in the subtask A and 49.47 (ranked 3/29) in the subtask B. The experimental results validate the effectiveness of our system.

pdf bib
THU_NGN at SemEval-2018 Task 1: Fine-grained Tweet Sentiment Intensity Analysis with Attention CNN-LSTM
Chuhan Wu | Fangzhao Wu | Junxin Liu | Zhigang Yuan | Sixing Wu | Yongfeng Huang
Proceedings of the 12th International Workshop on Semantic Evaluation

Traditional sentiment analysis approaches mainly focus on classifying the sentiment polarities or emotion categories of texts. However, they can’t exploit the sentiment intensity information. Therefore, the SemEval-2018 Task 1 is aimed to automatically determine the intensity of emotions or sentiment of tweets to mine fine-grained sentiment information. In order to address this task, we propose a system based on an attention CNN-LSTM model. In our model, LSTM is used to extract the long-term contextual information from texts. We apply attention techniques to selecting this information. A CNN layer with different size of kernels is used to extract local features. The dense layers take the pooled CNN feature maps and predict the intensity scores. Our system reaches average Pearson correlation score of 0.722 (ranked 12/48) in emotion intensity regression task, and 0.810 in valence regression task (ranked 15/38). It indicates that our system can be further extended.

pdf bib
THU_NGN at SemEval-2018 Task 2: Residual CNN-LSTM Network with Attention for English Emoji Prediction
Chuhan Wu | Fangzhao Wu | Sixing Wu | Zhigang Yuan | Junxin Liu | Yongfeng Huang
Proceedings of the 12th International Workshop on Semantic Evaluation

Emojis are widely used by social media and social network users when posting their messages. It is important to study the relationships between messages and emojis. Thus, in SemEval-2018 Task 2 an interesting and challenging task is proposed, i.e., predicting which emojis are evoked by text-based tweets. We propose a residual CNN-LSTM with attention (RCLA) model for this task. Our model combines CNN and LSTM layers to capture both local and long-range contextual information for tweet representation. In addition, attention mechanism is used to select important components. Besides, residual connection is applied to CNN layers to facilitate the training of neural networks. We also incorporated additional features such as POS tags and sentiment features extracted from lexicons. Our model achieved 30.25% macro-averaged F-score in the first subtask (i.e., emoji prediction in English), ranking 7th out of 48 participants.

pdf bib
THU_NGN at SemEval-2018 Task 10: Capturing Discriminative Attributes with MLP-CNN model
Chuhan Wu | Fangzhao Wu | Sixing Wu | Zhigang Yuan | Yongfeng Huang
Proceedings of the 12th International Workshop on Semantic Evaluation

Existing semantic models are capable of identifying the semantic similarity of words. However, it’s hard for these models to discriminate between a word and another similar word. Thus, the aim of SemEval-2018 Task 10 is to predict whether a word is a discriminative attribute between two concepts. In this task, we apply a multilayer perceptron (MLP)-convolutional neural network (CNN) model to identify whether an attribute is discriminative. The CNNs are used to extract low-level features from the inputs. The MLP takes both the flatten CNN maps and inputs to predict the labels. The evaluation F-score of our system on the test set is 0.629 (ranked 15th), which indicates that our system still needs to be improved. However, the behaviours of our system in our experiments provide useful information, which can help to improve the collective understanding of this novel task.

pdf bib
Neural Metaphor Detecting with CNN-LSTM Model
Chuhan Wu | Fangzhao Wu | Yubo Chen | Sixing Wu | Zhigang Yuan | Yongfeng Huang
Proceedings of the Workshop on Figurative Language Processing

Metaphors are figurative languages widely used in daily life and literatures. It’s an important task to detect the metaphors evoked by texts. Thus, the metaphor shared task is aimed to extract metaphors from plain texts at word level. We propose to use a CNN-LSTM model for this task. Our model combines CNN and LSTM layers to utilize both local and long-range contextual information for identifying metaphorical information. In addition, we compare the performance of the softmax classifier and conditional random field (CRF) for sequential labeling in this task. We also incorporated some additional features such as part of speech (POS) tags and word cluster to improve the performance of model. Our best model achieved 65.06% F-score in the all POS testing subtask and 67.15% in the verbs testing subtask.

pdf bib
Detecting Tweets Mentioning Drug Name and Adverse Drug Reaction with Hierarchical Tweet Representation and Multi-Head Self-Attention
Chuhan Wu | Fangzhao Wu | Junxin Liu | Sixing Wu | Yongfeng Huang | Xing Xie
Proceedings of the 2018 EMNLP Workshop SMM4H: The 3rd Social Media Mining for Health Applications Workshop & Shared Task

This paper describes our system for the first and third shared tasks of the third Social Media Mining for Health Applications (SMM4H) workshop, which aims to detect the tweets mentioning drug names and adverse drug reactions. In our system we propose a neural approach with hierarchical tweet representation and multi-head self-attention (HTR-MSA) for both tasks. Our system achieved the first place in both the first and third shared tasks of SMM4H with an F-score of 91.83% and 52.20% respectively.


pdf bib
THU_NGN at IJCNLP-2017 Task 2: Dimensional Sentiment Analysis for Chinese Phrases with Deep LSTM
Chuhan Wu | Fangzhao Wu | Yongfeng Huang | Sixing Wu | Zhigang Yuan
Proceedings of the IJCNLP 2017, Shared Tasks

Predicting valence-arousal ratings for words and phrases is very useful for constructing affective resources for dimensional sentiment analysis. Since the existing valence-arousal resources of Chinese are mainly in word-level and there is a lack of phrase-level ones, the Dimensional Sentiment Analysis for Chinese Phrases (DSAP) task aims to predict the valence-arousal ratings for Chinese affective words and phrases automatically. In this task, we propose an approach using a densely connected LSTM network and word features to identify dimensional sentiment on valence and arousal for words and phrases jointly. We use word embedding as major feature and choose part of speech (POS) and word clusters as additional features to train the dense LSTM network. The evaluation results of our submissions (1st and 2nd in average performance) validate the effectiveness of our system to predict valence and arousal dimensions for Chinese words and phrases.