Data exploration | EPFL Graph Search

Data exploration is an approach similar to initial data analysis, whereby a data analyst uses visual exploration to understand what is in a dataset and the characteristics of the data, rather than through traditional data management systems. These characteristics can include size or amount of data, completeness of the data, correctness of the data, possible relationships amongst data elements or files/tables in the data. Data exploration is typically conducted using a combination of automated and manual activities. Automated activities can include data profiling or data visualization or tabular reports to give the analyst an initial view into the data and an understanding of key characteristics. This is often followed by manual drill-down or filtering of the data to identify anomalies or patterns identified through the automated actions. Data exploration can also require manual scripting and queries into the data (e.g. using languages such as

Alpine: Efficient In-Situ Data Exploration in the Presence of Updates

Anastasia Ailamaki, Antonios Anagnostou, Matthaios Alexandros Olma

The ever growing data collections create the need for brief explorations of the available data to extract relevant information before decision making becomes necessary. In this context of data exploration, current data analysis solutions struggle to quickly pinpoint useful information in data collections. One major reason is that loading data in a DBMS without knowing which part of it will actually be useful is a major bottleneck. To remove this bottleneck, state-of-the art approaches perform queries in situ, thus avoiding the loading overhead. In situ query engines, however, are index-oblivious, and lack sophisticated techniques to reduce the amount of data to be accessed. Furthermore, applications constantly generate fresh data and update the existing raw data files whereas state-of-the art in situ approaches support only append-like workloads. In this demonstration, we showcase the efficiency of adaptive indexing and partitioning techniques for analytical queries in the presence of updates. We demonstrate an online partitioning and indexing tuner for in situ querying which plugs to a query engine and offers support for fast queries over raw data files. We present Alpine, our prototype implementation, which combines the tuner with a query executor incorporating in situ query techniques to provide efficient raw data access. We will visually demonstrate how Alpine incrementally and adaptively builds auxiliary data structures and indexes over raw data files and how it adapts its behavior as a side-effect of updates in the raw data files.

ACM New York, NY, USA2017

Efficient Query Processing for Spatial and Temporal Data Exploration

Eleni Tzirita Zacharatou

Core to many scientific and analytics applications are spatial data capturing the position or shape of objects in space, and time series recording the values of a process over time. Effective analysis of such data requires a shift from confirmatory pipelines to exploratory ones. However, there is a mismatch between the requirements of spatial and temporal data exploration and the capabilities of the data management solutions available today. First, traditional spatial query operators evaluate spatial relations with time-consuming geometric tests that oppose the interactivity expected from exploratory applications, creating an undue overhead. Second, spatial access methods are built on top of rough geometric object approximations that do not capture the complex structure and distribution of today's spatial data and are thus inefficient. Third, traditional indexes are typically built upfront before queries can be processed and over single data attributes, thus precluding interactive accesses to interesting data subsets that may be specified with constraints on multiple attributes. Finally, existing access methods scale poorly with increasingly granular spatial and temporal data originating from ever more precise data acquisition technologies and ever faster computing infrastructure. This thesis introduces a novel family of spatial and temporal access methods and query operators that aim to bridge the gap between existing data management techniques and data exploration applications. We show that spatial query operators can be decomposed into primitive graphics operations that are efficiently executed by graphics hardware (GPU) and allow to trade accuracy for interactivity. Furthermore, we design access methods that adapt to data characteristics, data growth trends, and workload access patterns, thereby providing scalable performance for ad-hoc queries over increasing data amounts. Specifically, we introduce a spatial approximation that adapts to the structural properties and distribution of the data, and propose spatial and time series access methods that leverage similarities between data items and support filtering over multiple attributes. Finally, we present an approach that indexes data incrementally, using queries as hints for optimizing data access.

EPFL2019

GPU Rasterization for Real-Time Spatial Aggregation over Arbitrary Polygons

Anastasia Ailamaki, Eleni Tzirita Zacharatou

Visual exploration of spatial data relies heavily on spatial aggregation queries that slice and summarize the data over different regions. These queries comprise computationally-intensive point-in-polygon tests that associate data points to polygonal regions, challenging the responsiveness of visualization tools. This challenge is compounded by the sheer amounts of data, requiring a large number of such tests to be performed. Traditional pre-aggregation approaches are unsuitable in this setting since they fix the query constraints and support only rectangular regions. On the other hand, query constraints are defined interactively in visual analytics systems, and polygons can be of arbitrary shapes. In this paper, we convert a spatial aggregation query into a set of drawing operations on a canvas and leverage the rendering pipeline of the graphics hardware (GPU) to enable interactive response times. Our technique trades-off accuracy for response time by adjusting the canvas resolution, and can even provide accurate results when combined with a polygon index. We evaluate our technique on two large real-world data sets, exhibiting superior performance compared to index-based approaches.

VLDB Endowment Inc.2017

Efficient Query Processing for Spatial and Temporal Data Exploration

Eleni Tzirita Zacharatou

EPFL2019