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

Background Based Conversations (BBCs) have been introduced to helpconversational systems avoid generating overly generic responses. In a BBC, theconversation is grounded in a knowledge source. A key challenge in BBCs isKnowledge Selection (KS): given a conversational context, try to find theappropriate background knowledge (a text fragment containing related facts orcomments, etc.) based on which to generate the next response. Previous workaddresses KS by employing attention and/or pointer mechanisms. These mechanismsuse a local perspective, i.e., they select a token at a time based solely onthe current decoding state. We argue for the adoption of a global perspective,i.e., pre-selecting some text fragments from the background knowledge thatcould help determine the topic of the next response. We enhance KS in BBCs byintroducing a Global-to-Local Knowledge Selection (GLKS) mechanism. Given aconversational context and background knowledge, we first learn a topictransition vector to encode the most likely text fragments to be used in thenext response, which is then used to guide the local KS at each decodingtimestamp. In order to effectively learn the topic transition vector, wepropose a distantly supervised learning schema. Experimental results show thatthe GLKS model significantly outperforms state-of-the-art methods in terms ofboth automatic and human evaluation. More importantly, GLKS achieves thiswithout requiring any extra annotations, which demonstrates its high degree ofscalability.

Thinking Globally, Acting Locally: Distantly Supervised Global-to-Local Knowledge Selection for Background Based Conversation