Differences

This shows you the differences between two versions of the page.

--- sql-guessing-advantage-analyser [2019/09/26 15:18]
alisa [Quick guide]
+++ sql-guessing-advantage-analyser [2021/06/14 11:46] (current)
alisa [Quick guide]
@@ Line 9: / Line 9: @@
 ===== Quick guide =====
-Compared to [[sql-derivative-sensitivity-analyser|combined sensitivity analyser]], the data objects of a model also have schemas and data tables, but now there are no explicit table norms. The distance measure for differential privacy will be determined in a different way.
+Similarly to [[sql-derivative-sensitivity-analyser|combined sensitivity analyser]], the data objects of a model also have schemas and data tables. However, there are no explicit table norms. The distance measure for differential privacy will be determined in a different way.
-Clicking on //Analyze// button opens a menu entitled //Analysis settings// on the right side of the page (in sidebar). In addition to the error level confidence slider that we have in [[sql-derivative-sensitivity-analyser|combined sensitivity analyser]], there is another slider allows to set desired upper bound on attacker’s advantage, which ranges between 0% and 100%.
+=== Table constraints ===
-{{slider2.png}}
+The syntax for table constraints is similar to [[sql-derivative-sensitivity-analyser|combined sensitivity analyser]], with some extensions. The keyword ''exact'' states that the attacker already knows some attribute precisely, and ''total'' says how many elements there can be, without specifying their values. More options can be found in [[sql-derivative-sensitivity-analyser_advanced|analyser advanced settings]]. Differently from the sensitivity analyser, the constraints are specified in a tab //Table Constraints// that opens after clicking on a table object, and the attributes are specified without table name as a prefix.
-The user has to specify the attacker's goal. Similarly to [[sql-derivative-sensitivity-analyser|combined sensitivity analyser]], user can define constraints on table attributes that are known in advance. There are two buttons for this.
-=== Table constraints ===
-The syntax for table constraints is the same as for [[sql-derivative-sensitivity-analyser|combined sensitivity analyser]], with some extensions. The keywords ''total'', ''set'' and ''range'' do not specify any probability distribution on the data, and the analyser assumes worst-case distribution by default (i.e. one for which the advantage is the largest). The keywords ''totalUnif'', ''setunif'', ''rangeUnif'' in addition specify that the distribution is uniform.
 <code>
-table_1.attr_1 exact;               --attacker knows the exact value
+attr_1 exact;               --attacker knows the exact value
+attr_2 total int;           --there are n possible values
+attr_3 set v1 ... vn;       --there are values {v1 ... vn}
+attr_4 range lb ub;         --the values come from range [lb,ub)
+</code>
-table_2.attr_2 total int;           --there are n possible values
+=== Attacker goal ===
-table_3.attr_3 set v1 ... vn;       --there are values {v1 ... vn}
-table_4.attr_4 range lb ub;         --the values come from range [lb,ub)
-table_5.attr_5 totalUnif int;       -- there are n uniformly distributed values
+Clicking on //Guessing Advantage analysis// button opens a menu entitled //Analysis settings// on the right side of the page (in sidebar). In addition to the error level confidence slider that we have in [[sql-derivative-sensitivity-analyser|combined sensitivity analyser]], there is another slider allows to set desired upper bound on attacker’s advantage, which ranges between 0% and 100%.
-table_6.attr_6 setUnif v1 ... vn;   -- uniformly distributed in {v1 ... vn}
-table_7.attr_7 rangeUnif lb ub;     -- uniformly distributed in [lb,ub)
--- value vk comes with probability pk
+{{slider2-v2.png}}
-table_8.attr_8 setPrior (v1, p1) ... (vn, pn)
--- range [v(k-1)...vk) comes with prob.pk
--- the values within [v(k-1)...vk) are distributed uniformly
-table_9.attr_9 rangePrior v0 (v1, p1) ... (vn, pn)
-</code>
-=== Attacker goal ===
-Attacker goal is given in form of an SQL guery. It defines a set of sensitive components, which the attacker is trying to guess. For each sensitive attribute, the guess can either be ''exact'' (discrete attributes), or ''approx r'' (approximated by r > 0 units). It is possible to combine several attributes into a vector and define approximation w.r.t. some l_p-norm as ''approxWrtLp(p)'' and ''approxWrtLinf''. The guesses can be combined into an expression using AND and OR operation, describing the case where leakage is considered successful. The filter ''WHERE condition'' describes which rows of the considered tables are treated as sensitive. The delimiter '';'' finishes the description of the attacker goal.
+Clicking the button //Attacker goal// opens a text window for specifying attacker's goal. The goal is given in form of an SQL guery, and it defines a set of sensitive components, which the attacker is trying to guess. Even if the attacker cannot guess the location precisely, it can still be bad even if he guesses the location precisely enough, so we need to introduce approximation. For each sensitive attribute, the guess can either be ''exact'' (discrete attributes), or ''approx r'' (approximated by r > 0 units). The delimiter '';'' finishes the description of the attacker goal.
 <code>
 SELECT
-(t.x, t.y) approxWrt(5) AND
+t.x approx 5 AND
-t.z exact
+t.y approx 5
-FROM t
+FROM t;
-WHERE t.b;
 </code>
-In this example, the attacker wins iff he guesses //both// ''t.z'' exactly and ''(t.x,t.y)'' within 5-unit precision w.r.t. l_2-norm of any row of the table ''t'' where ''t.b'' holds. The definition of "unit" depends on the data table, e.g. if the location was defined in miles, then a unit is also a mile.
+In this example, the attacker wins iff he guesses both ''t.x'' and ''t.y'' within 5-unit precision. The definition of "unit" depends on the data table, e.g. if the location was defined in miles, then a unit is also a mile.
-If we want to express that the attacker wins if he guesses //either// ''t.z'' or ''(t.x,t.y)'', we replace AND operation with OR.
+Additional syntax for specifying attacker goal can be found in [[sql-derivative-sensitivity-analyser_advanced|analyser advanced settings]].
 === Running analysis ===
-Click on //Run analysis// button to run analysis. The analyser internally converts these values to a suitable ε for differential privacy, and computes the noise required to achieve the bound on attacker’s advantage. The results (entitled //Analysis results//) appear in the sidebar as well. The result consists of the following components.
+Select the tasks that should be considered for the analysis (E.g. all the data processing tasks up to the computations of the value that the adversary uses for the guess). You should run the propagation if the part of the workflow that is analyzed does not start from the input objects of the process.
+Click on //Run analysis// button to run analysis. The analyser internally converts these values to a suitable ε for differential privacy, and computes the noise required to achieve the bound on attacker’s advantage. The results (entitled //Analysis results//) appear in the sidebar as well. The result consists of the following components, which are the same as for [[sql-derivative-sensitivity-analyser|combined sensitivity analyser]].
   * **actual outputs y** are the true outputs of the query, without noise.
@@ Line 61: / Line 48: @@
   * **p%-realtive error |a|/|y|** is the quotient of the additive noise and the query output. If there are several outputs, it is the quotient of corresponding vector norms.
-To see more precise values of prior and posterior guessing probabities, click //View more//. This can be useful for choosing appropriate value on the guessing advantage slider. For example, if the prior guessing probability was already 75%, then any value above 25% makes no sense since it would mean that the attacker is allowed to learn everything. Clicking //View more// also provides more information about how the noise should actually be generated, and it does it for Cauchy and Laplace noise distributions.
+To see more precise values of prior and posterior guessing probabilities, click //View more//. This can be useful for choosing appropriate value on the guessing advantage slider. For example, if the prior guessing probability was already 75%, then any value above 25% makes no sense since it would mean that the attacker is allowed to learn everything. Clicking //View more// also provides more information about how the noise should actually be generated, and it does it for Cauchy and Laplace noise distributions.
 ===== Source code =====
 The source code of SQL guessing advantage editor is available at [[https://github.com/pleak-tools/pleak-guessing-advantage-editor|pleak-sql-guessing-advantage-editor]] and the source code of SQL sensitivity analysis tools at [[https://github.com/pleak-tools/pleak-sql-analysis|pleak-sql-analysis]] repositories. Installation details can be found at [[sql-derivative-sensitivity-analyser_install|analyser installation guide]].

PLEAK Wiki

User Tools

Site Tools

Differences

Page Tools