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

With the prevalence of live broadcast business nowadays, a new type ofrecommendation service, called live broadcast recommendation, is widely used inmany mobile e-commerce Apps. Different from classical item recommendation, livebroadcast recommendation is to automatically recommend user anchors instead ofitems considering the interactions among triple-objects (i.e., users, anchors,items) rather than binary interactions between users and items. Existingmethods based on binary objects, ranging from early matrix factorization torecently emerged deep learning, obtain objects' embeddings by mapping frompre-existing features. Directly applying these techniques would lead to limitedperformance, as they are failing to encode collaborative signals amongtriple-objects. In this paper, we propose a novel TWo-side Interactive NetworkS(TWINS) for live broadcast recommendation. In order to fully use both staticand dynamic information on user and anchor sides, we combine a product-basedneural network with a recurrent neural network to learn the embedding of eachobject. In addition, instead of directly measuring the similarity, TWINSeffectively injects the collaborative effects into the embedding process in anexplicit manner by modeling interactive patterns between the user's browsinghistory and the anchor's broadcast history in both item and anchor aspects.Furthermore, we design a novel co-retrieval technique to select key items amongmassive historic records efficiently. Offline experiments on real large-scaledata show the superior performance of the proposed TWINS, compared torepresentative methods; and further results of online experiments on DiantaoApp show that TWINS gains average performance improvement of around 8% on ACTRmetric, 3% on UCTR metric, 3.5% on UCVR metric.

Who to Watch Next: Two-side Interactive Networks for Live Broadcast Recommendation