• Publications
  • Influence
DailyDialog: A Manually Labelled Multi-turn Dialogue Dataset
TLDR
A high-quality multi-turn dialog dataset, DailyDialog, which is intriguing in several aspects, the language is human-written and less noisy and the dialogues reflect the authors' daily communication way and cover various topics about their daily life. Expand
RippleNet: Propagating User Preferences on the Knowledge Graph for Recommender Systems
TLDR
This paper proposes RippleNet, an end-to-end framework that naturally incorporates the knowledge graph into recommender systems, and demonstrates that RippleNet achieves substantial gains in a variety of scenarios, including movie, book and news recommendation, over several state-of-the-art baselines. Expand
Knowledge Graph Convolutional Networks for Recommender Systems
TLDR
This paper proposes Knowledge Graph Convolutional Networks (KGCN), an end-to-end framework that captures inter-item relatedness effectively by mining their associated attributes on the KG. Expand
Knowledge-aware Graph Neural Networks with Label Smoothness Regularization for Recommender Systems
TLDR
This work proposes Knowledge-aware Graph Neural Networks with Label Smoothness regularization (KGNN-LS), which relies on label smoothness assumption, which posits that adjacent items in the knowledge graph are likely to have similar user relevance labels/scores, and proves that it is equivalent to a label propagation scheme on a graph. Expand
Applying regression models to query-focused multi-document summarization
TLDR
This paper presents a different kind of learning models, namely regression models, to query-focused multi-document summarization, and chooses to use Support Vector Regression to estimate the importance of a sentence in a document set to be summarized through a set of pre-defined features. Expand
Faithful to the Original: Fact Aware Neural Abstractive Summarization
TLDR
This work argues that faithfulness is also a vital prerequisite for a practical abstractive summarization system and proposes a dual-attention sequence-to-sequence framework to force the generation conditioned on both the source text and the extracted fact descriptions. Expand
Text-level Discourse Dependency Parsing
TLDR
The state-of-the-art dependency parsing techniques, the Eisner algorithm and maximum spanning tree (MST) algorithm, are adopted to parse an optimal discourse dependency tree based on the arcfactored model and the large-margin learning techniques. Expand
Unpaired Sentiment-to-Sentiment Translation: A Cycled Reinforcement Learning Approach
TLDR
A cycled reinforcement learning method that enables training on unpaired data by collaboration between a neutralization module and an emotionalization module that significantly outperforms the state-of-the-art systems. Expand
Multi-Task Feature Learning for Knowledge Graph Enhanced Recommendation
TLDR
This paper considers knowledge graphs as the source of side information and proposes MKR, a Multi-task feature learning approach for Knowledge graph enhanced Recommendation, a deep end-to-end framework that utilizes knowledge graph embedding task to assist recommendation task. Expand
Learning Summary Prior Representation for Extractive Summarization
TLDR
A novel summary system called PriorSum is developed, which applies the enhanced convolutional neural networks to capture the summary prior features derived from length-variable phrases under a regression framework, and concatenated with document-dependent features for sentence ranking. Expand
...
1
2
3
4
5
...