Method and computer-program-product determining measures for the development, design and/or deployment of complex embedded or cyber-physical systems, in particular complex software architectures used therein, of different technical domains

Patent metadata

Field	Value
Publication number	US-11294669-B2
Application number	US-201916960252-A
Country	US
Kind code	B2
Filing date	Mar 4, 2019
Priority date	Mar 5, 2018
Publication date	Apr 5, 2022
Grant date	Apr 5, 2022

Abstract

In order to improve the development, design and/or deployment of the complex embedded or the cyber-physical systems and in particular the complex software architectures being included and/or used therein of the different technical domains it is proposed a sophisticated expertise or tool-significant Measure Recommendation System. This Measure Recommendation System, particularly formed by the Computer-Program-Product, is tailored to determine and provide appropriate measures to effectively improve the decision-making process during the design, development and/or deployment of the systems of the different technical domains by automatically providing Measure Determining Vectors.

First claim

The invention claimed is:

1. A method for determining measures for at least one of the development, design and deployment of complex embedded or cyber-physical systems, wherein complex software architectures are used therein, of different technical domains, comprising:
a) using an issue Evaluating Phase, in which
a1) resolved issues of an Issue Tracking Database System either already structured or unstructured, and based on different kinds of data sources, which contain each at least one identification data field concerning at least a “Resolver”-attribute of a respective issue and one content related data field concerning at least “Summary”- and/or “Description”-attributes of the respective issue, are automatically edited and evaluated such that
the resolved issues are normalized and accordingly stored as normalized resolved issues in a Knowledge Database System,
a2) decisions being inherent in the normalized resolved issues are identified and classified into categories by a machine learning algorithm,
a3) the normalized resolved issues being identified and classified regarding the decisions are provided with labels,
a4) the content related data field of the normalized, decision-labeled issues are analyzed such that
a41) on the basis of general data of an Ontology Database System, Relevant Elements are determined,
a42) the normalized, decision-labeled issues, the Relevant Elements have been determined for, are provided with markers,

a5) the normalized, decision-labeled, element-marked issues stored in the Knowledge Database System are processed such that
a51) an Issue Evaluation Matrix is created, wherein
a511) matrix elements of the Issue Evaluation Matrix are formed on the one hand by the “Resolver”-attributes in the identification data fields of the normalized, decision-labeled, element-marked issues and on the other hand by the Relevant Elements determined for the normalized, decision-labeled, element-marked issues and
a512) each matrix element is weighted by determining the relevance, of the normalized, decision-labeled, element-marked issues having for the said matrix element the same matrix element constellation,

b) using a Measure Determining Phase, in which
b1) at least one unresolved issue of the Issue Tracking Database System either already structured or unstructured, regarding type and format, and based on one of the different kinds of data sources, which contains each at least one content related data field concerning at least “Summary”- and/or “Description”-attributes of said issue, is automatically edited and evaluated such that the unresolved issue is normalized and accordingly stored as a normalized unresolved issue in the Knowledge Database System,
b2) at least one further decision being inherent to the normalized unresolved issue is identified and classified into the categories by a machine learning algorithm,
b3) the normalized unresolved issue being identified and classified regarding the further decision is provided with a further label,
b4) the content related data field of the normalized, further decision-labeled issue is analyzed such that
b41) on the basis of the general data of the Ontology Database System at least one further Relevant Element is determined,
b42) the normalized, further decision-labeled issue, the further relevant element has been determined for, is provided with a further marker,

b5) the normalized, further decision-labeled, further element-marked issue stored in the Knowledge Database System is processed such that
an Issue Evaluation Vector is created, wherein the vector elements of the Issue Evaluation Vector with a number of vector elements, which correspond to the number of Relevant Elements determined for the normalized, decision-labeled, element-marked issues in the Issue Evaluation Matrix, are formed by the further Relevant Element determined for the normalized, further decision-labeled, further element-marked issue, which correspond each to one Relevant Element of the Relevant Elements determined for the normalized, decision-labeled, element-marked issues in the Issue Evaluation Matrix, and filled with “zero-value”-elements accordingly, if there is no “Relevant Element-Correspondence”,
b6) a measure for the development, design and/or deployment of the complex embedded or the cyber-physical systems, of the different technical domains is determined on the basis of a Measure Determination Vector generated by vector products of each the Issue Evaluation Vector and a dedicated Resolver Profile Vector, made up each from those matrix elements of the Issue Evaluation Matrix with the same “Resolver”-attribute of a set of Resolver Profile Vectors in the Issue Evaluation Matrix.

2. The method according to claim 1 , wherein the “Resolver” attributes are experts, given by their names, regarding at least one of the development, design and deployment of the complex embedded or the cyber-physical systems of the different technical domains.

3. The method according to claim 1 , wherein the “Resolver” attributes are tools regarding at least one of the development, design and deployment of the complex embedded or the cyber-physical systems, wherein the complex software architectures used therein, of the different technical domains.

4. The method according to claim 2 , wherein a set of rows in the Issue Evaluation Matrix are denoted with the experts or tools and accordingly the Resolver Profile Vector of the set of the Resolver Profile Vectors becomes a Expert/Tool Profile Vector.

5. The method according to claim 2 , wherein an “expertList/toolList” is created by matching and prioritizing on the basis of the Measure Determination Vector being generated according to pseudo code described by the following algorithm:

//matching

 1:
function MATCH(IEV, M m,n , D)

 2:
“expertList/toolList” ← { }

 3:
for i in 0..m do

 4:
MDV ← newArray(n);

 5:
for j in 0..n do

 6:
MDV[... j ] ← IEV[... j ] x M[... i ,... j ], wherein M[... i ,... j ] =(Resolver

Profile Vector i )

 7:
end for

//prioritizing by computing a score as a vector magnitude

 8:
sum ← 0

 9:
for j in 0..n do

10:
sum ← sum + MDV[... j ] x MDV[... j ]

11:
end for

12:
score ← SQRT(sum)

13:
if score >0 then

14:
“expertList/toolList”:add(“person”/“tool”, D[... i ])

15:
“expertList/toolList” :add(“score”, score)

16:
end if

17:
end for

18:
“expertList/toolList” ← ORDERBY(“expertList/toolList”, “score”)

19:

Assignees

• Siemens Ag

Inventors

• Hohenstein Uwe
• Biesdorf Andreas
• Brand Christoph
• Kramer Martin
• Mahabaleshwar Manoj
• Shumaiev Klym

Classifications

• G06F8/70 Primary

  Software maintenance or management · CPC title

• G06F8/61

  Installation · CPC title

• G06F17/16

  Matrix or vector computation {, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization (matrix transposition G06F7/78)} · CPC title

• G06F8/77 Primary

  Software metrics · CPC title