English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Multi-party dialogues are more difficult for models to understand thanone-to-one two-party dialogues, since they involve multiple interlocutors,resulting in interweaving reply-to relations and information flows. To stepover these obstacles, an effective way is to pre-train a model that understandsthe discourse structure of multi-party dialogues, namely, to whom eachutterance is replying. However, due to the lack of explicitly annotateddiscourse labels in multi-party dialogue corpora, previous works fail to scaleup the pre-training process by putting aside the unlabeled multi-partyconversational data for nothing. To fully utilize the unlabeled data, wepropose to treat the discourse structures as latent variables, then jointlyinfer them and pre-train the discourse-aware model by unsupervised latentvariable inference methods. Experiments on multiple downstream tasks show thatour pre-trained model outperforms strong baselines by large margins andachieves state-of-the-art (SOTA) results, justifying the effectiveness of ourmethod. The official implementation of this paper is available athttps://github.com/EricLee8/MPD_EMVI.

Pre-training Multi-party Dialogue Models with Latent Discourse Inference