Querying data provenance

Abstract

Many advanced data management operations (e.g., incremental maintenance, trust assessment, debugging schema mappings, keyword search over databases, or query answering in probabilistic databases), involve computations that look at how a tuple was produced, e.g., to determine its score or existence. This requires answers to queries such as, "Is this data derivable from trusted tuples?"; "What tuples are derived from this relation?"; or "What score should this answer receive, given initial scores of the base tuples?". Such questions can be answered by consulting the provenance of query results.In recent years there has been significant progress on formal models for provenance. However, the issues of provenance storage, maintenance, and querying have not yet been addressed in an application-independent way. In this paper, we adopt the most general formalism for tuple-based provenance, semiring provenance. We develop a query language for provenance, which can express all of the aforementioned types of queries, as well as many more; we propose storage, processing and indexing schemes for data provenance in support of these queries; and we experimentally validate the feasibility of provenance querying and the benefits of our indexing techniques across a variety of application classes and queries.

A Query Language of Data Provenance Based on Dependency View for Process Analysis

“…However, the difficulty to support process analysis based on data provenance is mainly the complexity and variety of the analysis requirement, which proposes a serious description problem for data provenance querying. Then the issue of data provenance querying is just addressed in a application-independent way to in recent years, the query languages of data provenance are proposed to resolve the problem, like OPQL [6], VQuel [7], ProQL [8], QLP [9], etc. These query languages are based on formal models for data provenance, which cause the difficulty in writing query by these languages.…”

Fine-Grained Provenance of Users’ Interpretations in a Collaborative Visualization Architecture

“…Our data structure is fine-grained so that each voxel-equivalent data unit is stored as a tuple, a row in a database. To support querying of tuple-based provenance (Karvounarakis et al, 2010), we restructure our data into a relational form. As a result, our visual analysis method shifts from the classic static raw file system into a central data structure built on a relational database.…”

A Formal Framework for Provenance Security

“…For most of these models, there are semantic guarantees (or even exact semantic characterizations) relating the provenance records to the denotation of the program. However, even for the semiring model, basic questions such as query equivalence for annotated relations [30] and how to implement provenance and query provenance-annotated databases [31], [32] are only beginning to be addressed.…”