Raftul cu initiativa Book Archive


Preserving Privacy in On-Line Analytical Processing (OLAP) by Lingyu Wang

By Lingyu Wang

On-Line Analytic Processing (OLAP) platforms frequently have to meet conflicting ambitions. First, the delicate info kept in underlying info warehouses needs to be saved mystery. moment, analytical queries concerning the information needs to be allowed for selection help reasons. the most problem is that delicate information will be inferred from solutions to possible blameless aggregations of the knowledge. present inference keep an eye on equipment in statistical databases often convey excessive functionality overhead and constrained effectiveness whilst utilized to OLAP systems.

Preserving privateness in online Analytical Processing stories a chain of tools that may accurately resolution info cube-style OLAP queries relating to delicate info whereas provably fighting adversaries from inferring the knowledge. how you can maintain the functionality overhead of those safety tools at an affordable point is additionally addressed. reaching a stability among protection, availability, and function is proven to be possible in OLAP systems.

Preserving privateness in online Analytical Processing is designed for the pro industry, composed of practitioners and researchers in undefined. This publication can be acceptable for graduate-level scholars in laptop technology and engineering.

Show description

Read Online or Download Preserving Privacy in On-Line Analytical Processing (OLAP) PDF

Similar comptia books

Internet Security Cryptographic Principles Algorithms and Protocols

Wisdom of quantity conception and summary algebra are pre-requisites for any engineer designing a safe internet-based procedure. despite the fact that, many of the books at present on hand at the topic are geared toward practitioners who simply need to know how a few of the instruments in the marketplace paintings and what point of safeguard they convey.

CompTIA A+ 220-701 and 220-702 Cert Guide

Enable me commence through asserting that this present day I took the 220-701, the 1st of the necessary assessments. I handed with a ranking of 775, which as most sensible i will determine correlates to among eighty five and ninety on a a hundred element scale. My learn was once solely self-directed and consisted of utilizing 4 assorted books, the A+ video sequence from okay Alliance and examination prep software program from [.

Pro PHP Security (Pro)

Hypertext Preprocessor is the world’s hottest open resource net scripting language, put in on virtually 17 million domain names around the globe (www. Hypertext Preprocessor. net/usage. php). it's enjoyed through novices and embraced via complex clients. This e-book deals builders a whole advisor to taking either shielding and proactive safety ways inside their personal home page functions.

Secure Data Management in Decentralized Systems

The examine scope of database safety has extended vastly, a result of swift improvement of the worldwide inter-networked infrastructure. Databases are not any longer stand-alone structures which are basically available to inner clients of agencies. as an alternative, permitting selective entry from diversified protection domain names has develop into a needs to for lots of company practices.

Extra resources for Preserving Privacy in On-Line Analytical Processing (OLAP)

Example text

Cell suppression is used to protect census data released in statistical tables [21, 221. Cells containing sensitive COUNTS are first suppressed according to a given sensitivity criterion. Possible inferences of the suppressed cells are then detected and re-moved using linear (or integer) programming-based techniques. While such a detection method is effective for two-dimensional cases, it is intractable for three or more dimensional tables even of small sizes [22, 251. Partitioning first defines a partition on the set of sensitive data, it then restricts queries to aggregate only complete blocks in the partition [15, 811.

We use the same notation [tl,t2], where tl and tz are two tuples, to specify range queries as before. We assume SUM queries nt=l[l, 40 4 Inferences in Data Cubes and omit the aggregation function. The following example illustrates these concepts. 1. We rephrase the previous example in above notations. The two dimensions are [ I ,21 and [I,41. The Cartesian product [1,2] x [ I , 41 includes eight possible tuples of which only six appear in the core cuboid ( ( 1 ,I ) , (1,2),(1,3),(2,2),( 2 , 3 ) ,(2,4)).

The above result is essentially a precise model for inferences of unbounded real values using SUM-only queries. The result also leads to a method for checking whether a new query, taken together with queries answered before, will cause inferences. A straightforward but inefficient approach is to keep all answered queries and re-computing the RREF when each new query is received. For m queries on n values, the GaussJordan elimination takes time O(m2n). Considering that the elementary row operations on a matrix is associative, a better approach is to incrementally updates the RREF for each newly answered query.

Download PDF sample

Rated 4.93 of 5 – based on 42 votes