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

Data analytics over normalized databases typically requires computing andmaterializing expensive joins (wide-tables). Factorized query execution modelsexecution as message passing between relations in the join graph and pushesaggregations through joins to reduce intermediate result sizes. Although thisaccelerates query execution, it only optimizes a single wide-table query. Incontrast, wide-table analytics is usually interactive and users want to applydelta to the initial query structure. For instance, users want to slice, diceand drill-down dimensions, update part of the tables and join with new tablesfor enrichment. Such Wide-table Delta Analytics offers novel work-sharingopportunities. This work shows that carefully materializing messages duringquery execution can accelerate Wide-table Delta Analytics by >10^5x as comparedto factorized execution, and only incurs a constant factor overhead. The keychallenge is that messages are sensitive to the message passing ordering. Toaddress this challenge, we borrow the concept of calibration in probabilisticgraphical models to materialize sufficient messages to support any ordering. Wemanifest these ideas in the novel Calibrated Junction Hypertree (CJT) datastructure, which is fast to build, aggressively re-uses messages to acceleratefuture queries, and is incrementally maintainable under updates. We furthershow how CJTs benefit applications such as OLAP, query explanation, streamingdata, and data augmentation for ML. Our experiments evaluate three versions ofthe CJT that run in a single-threaded custom engine, on cloud DBs, and inPandas, and show 30x - 10^5x improvements over state-of-the-art factorizedexecution algorithms on the above applications.

Calibration: A Simple Trick for Wide-table Delta Analytics