Papers by Hayato Kobayashi
Pretraining Sentiment Classifiers with Unlabeled Dialog Data (P18-2)
Copied to clipboard
| Challenge: | Existing methods to train sentiment classifiers with unlabeled data are costly and time-consuming. |
| Approach: | They propose a conditional language model with unlabeled dialog data instead of a language model to pretrain sentiment classifiers. |
| Outcome: | The proposed strategy outperforms state-of-the-art methods with unlabeled dialog data and is simple but effective. |
A Case Study on Neural Headline Generation for Editing Support (N19-2)
Copied to clipboard
Kazuma Murao, Ken Kobayashi, Hayato Kobayashi, Taichi Yatsuka, Takeshi Masuyama, Tatsuru Higurashi, Yoshimune Tabuchi
| Challenge: | a news-aggregator is a website or mobile application that aggregates web content . dozens of professional editors manually create their headlines, which are much shorter than the original headlines. |
| Approach: | They propose a neural headline generation model that automatically generates short headlines from news articles. |
| Outcome: | The proposed model is deployed to an editing support tool and compares editors' behavior before and after the release. |
Frustratingly Easy Model Ensemble for Abstractive Summarization (D18-1)
Copied to clipboard
| Challenge: | Existing studies on compressing or distilling ensemble models have shown that they increase computational costs and reduce performance. |
| Approach: | They propose an unsupervised method that combines multiple models by selecting a majority-like output in post-processing. |
| Outcome: | The proposed method performs better than the current ensemble methods on a news-headline-generation task. |
Dataset Creation for Ranking Constructive News Comments (P19-1)
Copied to clipboard
| Challenge: | Existing studies on comment ranking use user feedback as a quality measure . however, this type of measurement has two drawbacks: (a) user feedback does not always satisfy the service provider's needs, such as to create a fair place; and (b) user input will be biased by where comments appear in a comment thread. |
| Approach: | They propose to evaluate quality of comments on the basis of constructiveness separately from user feedback. |
| Outcome: | The proposed model improves on 100K+ Japanese comments with constructiveness scores . it shows that C-scores are not always related to positive feedback . |
Extractive Headline Generation Based on Learning to Rank for Community Question Answering (C18-1)
Copied to clipboard
| Challenge: | Community question answering (CQA) forums do not always have appropriate headlines because of user-generated content. |
| Approach: | They propose an extractive headline generation method that extracts the most informative substring from each question as its headline. |
| Outcome: | The proposed method outperforms baselines including a prefix-based method . it uses the prefix of a question as its headline to create the most informative substring . |