Jiayu Song


2022

pdf
Identifying Moments of Change from Longitudinal User Text
Adam Tsakalidis | Federico Nanni | Anthony Hills | Jenny Chim | Jiayu Song | Maria Liakata
Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)

Identifying changes in individuals’ behaviour and mood, as observed via content shared on online platforms, is increasingly gaining importance. Most research to-date on this topic focuses on either: (a) identifying individuals at risk or with a certain mental health condition given a batch of posts or (b) providing equivalent labels at the post level. A disadvantage of such work is the lack of a strong temporal component and the inability to make longitudinal assessments following an individual’s trajectory and allowing timely interventions. Here we define a new task, that of identifying moments of change in individuals on the basis of their shared content online. The changes we consider are sudden shifts in mood (switches) or gradual mood progression (escalations). We have created detailed guidelines for capturing moments of change and a corpus of 500 manually annotated user timelines (18.7K posts). We have developed a variety of baseline models drawing inspiration from related tasks and show that the best performance is obtained through context aware sequential modelling. We also introduce new metrics for capturing rare events in temporal windows.

pdf
Unsupervised Opinion Summarisation in the Wasserstein Space
Jiayu Song | Iman Munire Bilal | Adam Tsakalidis | Rob Procter | Maria Liakata
Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing

Opinion summarisation synthesises opinions expressed in a group of documents discussingthe same topic to produce a single summary. Recent work has looked at opinion summarisation of clusters of social media posts. Such posts are noisy and have unpredictable structure, posing additional challenges for the construction of the summary distribution and the preservation of meaning compared to online reviews, which has been so far the focus on opinion summarisation. To address these challenges we present WassOS, an unsupervised abstractive summarization model which makesuse of the Wasserstein distance. A Variational Autoencoder is first used to obtain the distribution of documents/posts, and the summary distribution is obtained as the Wasserstein barycenter. We create separate disentangled latent semantic and syntactic representations of the summary, which are fed into a GRU decoder with a transformer layer to produce the final summary. Our experiments onmultiple datasets including reviews, Twitter clusters and Reddit threads show that WassOSalmost always outperforms the state-of-the-art on ROUGE metrics and consistently producesthe best summaries with respect to meaning preservation according to human evaluations.

2019

pdf
An Encoding Strategy Based Word-Character LSTM for Chinese NER
Wei Liu | Tongge Xu | Qinghua Xu | Jiayu Song | Yueran Zu
Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers)

A recently proposed lattice model has demonstrated that words in character sequence can provide rich word boundary information for character-based Chinese NER model. In this model, word information is integrated into a shortcut path between the start and the end characters of the word. However, the existence of shortcut path may cause the model to degenerate into a partial word-based model, which will suffer from word segmentation errors. Furthermore, the lattice model can not be trained in batches due to its DAG structure. In this paper, we propose a novel word-character LSTM(WC-LSTM) model to add word information into the start or the end character of the word, alleviating the influence of word segmentation errors while obtaining the word boundary information. Four different strategies are explored in our model to encode word information into a fixed-sized representation for efficient batch training. Experiments on benchmark datasets show that our proposed model outperforms other state-of-the-arts models.