Skip to content

Commit

Permalink
Update role mining xsd schemes
Browse files Browse the repository at this point in the history
  • Loading branch information
@tchrapovic
tchrapovic committed Jul 17, 2023
1 parent 2703fdf commit 16eb20b
Showing 1 changed file with 65 additions and 63 deletions.
128 changes: 65 additions & 63 deletions infra/schema/src/main/resources/xml/ns/public/common/common-mining-3.xsd
View file
Original file line number Diff line number Diff line change
Expand Up @@ -51,9 +51,9 @@
<xsd:element name="elements" type="xsd:string" minOccurs="0" maxOccurs="unbounded">
<xsd:annotation>
<xsd:documentation>
The main object that was analyzed. In the case of ROLE MODE, the list of role object
The main object that was analyzed. In the case of ROLE MODE, the list of role object (elements = roles)
IDs that make up the given group is included in the object. In the case of USER MODE,
the list of users object IDs that make up the given group is included in the object.
the list of users object (elements = users) IDs that make up the given group is included in the object.
</xsd:documentation>
</xsd:annotation>
</xsd:element>
Expand All @@ -63,8 +63,8 @@
<xsd:documentation>
<p>
The absolute value of the elements in the given group.
In the case of USER MODE, it is the count of users.
In the case of ROLE MODE, it is the count of roles.
In the case of USER MODE, it is the count of users. (elements = users)
In the case of ROLE MODE, it is the count of roles. (elements = roles)
</p>
</xsd:documentation>
<xsd:appinfo>
Expand All @@ -77,10 +77,10 @@
<xsd:annotation>
<xsd:documentation>
<p>
The absolute value of the elements (objects) according to which the "elements" were included
The absolute value of the points (objects) according to which the "elements" were included
in the given group.
In the case of USER MODE, it is the count of roles.
In the case of ROLE MODE, it is the count of users.
In the case of USER MODE, it is the count of roles. (roles = points, elements = users)
In the case of ROLE MODE, it is the count of users. (users = points, elements = roles)
</p>
</xsd:documentation>
<xsd:appinfo>
Expand Down Expand Up @@ -133,9 +133,9 @@
<xsd:annotation>
<xsd:documentation>
<p>
Average number of compared objects per element in a given group (RoleAnalysisCluster).
In the case of USER MODE, it is the average number of roles.
In the case of ROLE MODE, it is the average number of users.
Average number of compared objects (points) per element in a given group (RoleAnalysisCluster).
In the case of USER MODE, it is the average number of roles. (points = roles)
In the case of ROLE MODE, it is the average number of users. (points = users)
</p>
</xsd:documentation>
<xsd:appinfo>
Expand All @@ -148,9 +148,9 @@
<xsd:annotation>
<xsd:documentation>
<p>
Minimum number of compared objects among the element in a given group (RoleAnalysisCluster).
In the case of USER MODE, it is the minimum roles set size.
In the case of ROLE MODE, it is the minimum users set size.
Minimum number of compared objects (points) among the element in a given group (RoleAnalysisCluster).
In the case of USER MODE, it is the minimum roles set size. (points = roles)
In the case of ROLE MODE, it is the minimum users set size. (points = users)
</p>
</xsd:documentation>
<xsd:appinfo>
Expand All @@ -163,9 +163,9 @@
<xsd:annotation>
<xsd:documentation>
<p>
Maximum number of compared objects among the element in a given group (RoleAnalysisCluster).
In the case of USER MODE, it is the highest roles set size.
In the case of ROLE MODE, it is the highest users set size.
Maximum number of compared objects (points) among the element in a given group (RoleAnalysisCluster).
In the case of USER MODE, it is the highest roles set size. (points = roles)
In the case of ROLE MODE, it is the highest users set size. (points = users)
</p>
</xsd:documentation>
<xsd:appinfo>
Expand Down Expand Up @@ -201,9 +201,9 @@
<xsd:annotation>
<xsd:documentation>
<p>
The RoleAnalysisSession represents a session for role analysis, providing a container for
RoleAnalysisCluster objects. It serves as a parent object that contains references and basic
information about its child clusters (RoleAnalysisSession).
The RoleAnalysisSession represents a session for role analysis, providing a container for
RoleAnalysisCluster objects. It serves as a parent object that contains references and basic
information about its child clusters (RoleAnalysisSession).
</p>
</xsd:documentation>
<xsd:appinfo>
Expand All @@ -219,32 +219,32 @@
<xsd:annotation>
<xsd:documentation>
<p>
The options element within the RoleAnalysisSession class represents cluster options for role
analysis. It is loaded as a JSON string format, allowing the configuration settings and choices
for the clusters to be defined in a structured and flexible manner.
This JSON string can be easily parsed and processed to extract the specific
options required for the role analysis session. The following keys are currently used:

Cluster Options Query": Specifies the query or filter applied to the userType object.
assignThreshold": Minimum count of relation with compared object. USER MODE - min role assign.
mode": Determines the process mode for role analysis.
similarity": Similarity threshold used for clustering (Jaccard).
minIntersection": Minimum number of intersections required.
minGroup": Specifies the minimum size of a group (unique objects).

For default business role suggestion:
detectIntersection": Minimum number of intersections required.
detectMaxFrequency": Maximum allowed frequency.
detectMinFrequency": Minimum required frequency.
detectOccupancy": Minimum occupancy.

These keys represent specific configuration options within the options element,
and their corresponding values define the desired settings for the role analysis session.

Furthermore, the role analysis session allows for the possibility of updating these values.
If there is a need to modify or refine the cluster options during the analysis process, the values
associated with the keys can be updated accordingly. This flexibility enables fine-tuning of the
role analysis based on changing requirements or new insights gained during the analysis phase.
The options element within the RoleAnalysisSession class represents cluster options for role
analysis. It is loaded as a JSON string format, allowing the configuration settings and choices
for the clusters to be defined in a structured and flexible manner.
This JSON string can be easily parsed and processed to extract the specific
options required for the role analysis session. The following keys are currently used:

Cluster Options Query: Specifies the query or filter applied to the userType object.
assignThreshold: Minimum count of relation with compared object. USER MODE - min role assign.
mode: Determines the process mode for role analysis.
similarity: Similarity threshold used for clustering (Jaccard).
minIntersection: Minimum number of intersections required.
minGroup: Specifies the minimum size of a group (unique objects).

For default business role suggestion:
detectIntersection: Minimum number of intersections required.
detectMaxFrequency: Maximum allowed frequency.
detectMinFrequency: Minimum required frequency.
detectOccupancy: Minimum occupancy.

These keys represent specific configuration options within the options element,
and their corresponding values define the desired settings for the role analysis session.

Furthermore, the role analysis session allows for the possibility of updating these values.
If there is a need to modify or refine the cluster options during the analysis process, the values
associated with the keys can be updated accordingly. This flexibility enables fine-tuning of the
role analysis based on changing requirements or new insights gained during the analysis phase.
</p>
</xsd:documentation>
</xsd:annotation>
Expand All @@ -254,22 +254,24 @@
<xsd:annotation>
<xsd:documentation>
<p>
The mode element specifies the process mode for role analysis. It determines the specific approach
or methodology used during the analysis. The value of this element is a string that describes the
process mode.

We currently use an enum type to identify the mode (ROLE MODE, USER MODE)
For Role Analysis (ROLE MODE):
Role perspective is taken into account. The role mining process focuses on identifying
and analyzing roles based on their similarity and considering the perspective of roles. Objects
that meet the specified conditions, determined by the input parameters, are added to the data set
to facilitate role mining

For User Analysis (USER MODE):
User perspective is taken into account. The user mining process focuses on identifying and
analyzing users based on their similarity and considering the perspective of users. Objects that
meet the specified conditions, determined by the input parameters, are added to the data set to
facilitate role mining.
The mode element specifies the process mode for role analysis. It determines the specific approach
or methodology used during the analysis. The value of this element is a string that describes the
process mode.

We currently use an enum type to identify the mode (ROLE MODE, USER MODE)
For Role Analysis (ROLE MODE):
Role perspective is taken into account. The role mining process focuses on identifying
and analyzing roles based on their similarity and considering the perspective of roles
(vectors of users who have this role are compared). Objects
that meet the specified conditions, determined by the input parameters, are added to the data set
to facilitate role mining

For User Analysis (USER MODE):
User perspective is taken into account. The user mining process focuses on identifying and
analyzing users based on their similarity and considering the perspective of users
(the role vectors assigned to the given users are compared). Objects that
meet the specified conditions, determined by the input parameters, are added to the data set to
facilitate role mining.
</p>
</xsd:documentation>
<xsd:appinfo>
Expand All @@ -282,8 +284,8 @@
<xsd:annotation>
<xsd:documentation>
<p>
The roleAnalysisClusterRef element represents references to clusters
associated with the role analysis session.
The roleAnalysisClusterRef element represents references to clusters
associated with the role analysis session.
</p>
</xsd:documentation>
</xsd:annotation>
Expand Down

0 comments on commit 16eb20b

Please sign in to comment.