A scalable direct manipulation engine for position-aware presentational data management

With the explosion of data, large datasets become more common for data analysis. How- ever, existing analytic tools are lack of scalability and large-scale data management tools are lack of interactivity. A lot of data analysis tasks are based on the order of data, we are proposing the very first positional storage engine supporting persistence and maintenance of orders for large datasets and allow direct manipulation on orders. We introduce a sparse monotonic order statistic structure for persisting and maintaining order. We also show how to support multiple orders and optimize the storage. After that, we demonstrate a buffered storage manager to ensure the direct manipulation interactivity. Last, we show our final system DataSpread which is interactive and scalable. In the end, we hope that our solution can point out a potential direction to support data analysis for large-scale data.

Name thesis:degree_name

M.S.

Level thesis:degree_level

Thesis

Discipline thesis:degree_discipline

Computer Science

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2018

Author dc:creator

• Zhou, Xinyan

Contributors dc:contributor

• Chang, Kevin Chen-Chuan

• Order
• Direct Manipulation
• Positional Indexing

Statement dc:rights

• Copyright 2018 Xinyan Zhou

Language dc:language

en

Handle dc:identifier

http://hdl.handle.net/2142/101629

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/101629