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/

Zendy Plus

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

Zendy Tools

Zendy Open

Identifying the topic (domain) of each user's utterance in open-domainconversational systems is a crucial step for all subsequent languageunderstanding and response tasks. In particular, for complex domains, anutterance is often routed to a single component responsible for that domain.Thus, correctly mapping a user utterance to the right domain is critical. Toaddress this problem, we introduce ConCET: a Concurrent Entity-awareconversational Topic classifier, which incorporates entity-type informationtogether with the utterance content features. Specifically, ConCET utilizesentity information to enrich the utterance representation, combining character,word, and entity-type embeddings into a single representation. However, forrich domains with millions of available entities, unrealistic amounts oflabeled training data would be required. To complement our model, we propose asimple and effective method for generating synthetic training data, to augmentthe typically limited amounts of labeled training data, using commonlyavailable knowledge bases to generate additional labeled utterances. Weextensively evaluate ConCET and our proposed training method first on an openlyavailable human-human conversational dataset called Self-Dialogue, to calibrateour approach against previous state-of-the-art methods; second, we evaluateConCET on a large dataset of human-machine conversations with real users,collected as part of the Amazon Alexa Prize. Our results show that ConCETsignificantly improves topic classification performance on both datasets,including 8-10% improvements over state-of-the-art deep learning methods. Wecomplement our quantitative results with detailed analysis of systemperformance, which could be used for further improvements of conversationalagents.

ConCET: Entity-Aware Topic Classification for Open-Domain Conversational Agents