SBVR2UML: A Challenging Transformation.

Full text

Page 1

SBVR2UML: A Challenging Transformation

Hina Afreen
Department of Computer Science & IT
The Islamia University of Bahawalpur
Bahawalpur, Pakistan
hina.afreen@hotmail.com
Imran Sarwar Bajwa, Behzad Bordbar
School of Computer Science & IT
University of Birmingham
Birmingham, UK
i.s.bajwa@cs.bham.ac.uk


Abstract— UML is a de-facto standard used for generating the
software models. UML support visualization of the software
artifacts. To generate a UML diagram, a software engineer has to
collect software requirements in a natural language (such as
English) or a semi-formal language (such as SBVR), manually
analyze the requirements and then manually generate the class
diagrams in an available CASE tool. However, by automatically
transforming SBVR Software requirements to UML can
seriously share burden of a system analyst and can improve the
quality and robustness of software modeling phase. The paper
demonstrates the challenging aspect of model transformation
from SBVR to UML. The presented approach takes input the
software requirements specified in SBVR syntax, parses the input
specification, extracts the UML ingredients such as classes,
methods, attributes, associations, etc and finally generate the
visual representation of the extracted information. The presented
approach is fully automated. The presented approach is
explained via an example.
Keywords- Automated Software Modelling, UML, SBVR
I.
INTRODUCTION
The emergence of Object Oriented Analysis and Design in
software engineering has led to the semi and fully automation
of various phases of software modeling. Automated generation
of Unified Modeling Language (UML) [1] diagrams from the
natural language software specifications is a major break
through in the field of automated software modeling [2], [3].
Examples of such work are LIDA [4], GOOAL [4], CM-
Builder [6], Re-Builder [7], NL-OOML [8], UML-Generator
[9], etc. However, most of the tools are at their preliminary
stages and do not provide very high accuracy [6]. A primary
reason attributed by the researchers is the ambiguous nature of
English that makes machine processing intricate and complex.
Where the trends in software modeling and software
programming are rapidly changing, major advancements have
been taken place in the field of software requirements
elicitation and specification [10]. One of the recent
advancement is use of OMG‟s is new standard SBVR
(Semantic of Business Vocabulary and Rules) [11] in
specification of the software requirements. The use of SBVR
in capturing the software requirements not only advances the
process of software specification but also simplifies the
process of analyzing and designing the software models.
Moreover, SBVR is based on higher order logic and hence
easy to machine process. We want to exploit these salient
features of SBVR and aim to machine processing SBVR and
transform SBVR to UML class models using the model
transformation technology.
A real challenge in SBVR to UML class model
transformation was to deal with the un-addressed issue in
available approaches for SBVR to UML transformations. For
example, the approaches presented by Raj [12] and Nemuraite
[13] do not support the automated parsing of SBVR
specifications. A user should have the pre-parsed SBVR
specification to use these approaches. Another drawback of
these approaches is that they perform partial mapping in
SBVR and UML class model. These approaches also lack
support for extracting class associations, aggregations and
generalizations [12], [13]. We aim to address these issues and
present a tool that can automatically parse SBVR and can
perform complete mapping from SBVR to UML.
In this paper, the major contribution is threefold. Firstly,
model transformation based a novel approach is presented to
perform syntactic and semantic analysis of SBVR
specification of software requirements to extract object
oriented elements as classes,
associations, generalizations, etc. Secondly, we report the
structure of the implemented tool SBVR2UML that is able to
automatically generate UML class models from SBVR
software requirements specifications. Thirdly, we have solved
a case study with our tool and compared the results with other
tools (used for automated OOA) for the sake of performance
evaluation.
The remaining paper is structured into the following
sections: Section 2 describes background and the related work.
Section 3 illustrates the architecture and workflow of the
presented tool, SBVR2UML. Section 4 presents a solved case
study from the domain of library information systems. Finally,
the paper is concluded to discuss the future work.
attributes, operations,
II.
BACKGROUND
In this section, a brief introduction to the basic concepts of
the MDA, UML and SBVR is provided. The elaborated
concepts are used in the rest of the paper.
A. Model Driven Architecture
The presented approach to model transform input SBVR

Page 2

specification to UML specification is based on the Model
Driven Architecture (MDA) [14]. MDA is a software design
approach typically involved in model based software system
development. A key concept in the MDA is the notion of
metamodels [15]. Model transformation can be defined in the
MDA by mapping elements of the source metamodel to the
target metamodel. On the basis of these mappings a set of
rules are defined, typically called Transformation Rules. By
employing these transformation rules, every model, which is
an instance of the source metamodel, can be automatically
transformed to an instance of the destination metamodel [16].
B. UMLClass Models
The Unified Modeling Language (UML) [1] supports a set
of diagrams to visually represent various software artifacts.
One of the most common and widely used diagrams is the
UML class diagram. The class diagrams are typically used to
represent the structural information of a software model. The
key element in a class model is class. A class can have sub-
elements such as attributes, methods, and associations.
C. SBVR Specification
A typical SBVR representation is based on SBVR business
vocabulary and SBVR business rules [11]. In SBVR, all the
specific terms and definitions of concepts used by an
organization or community in course of business are treated as
vocabulary. Common examples of SBVR vocabulary are
object types, individual concepts, characteristics, fact types,
etc. In SBVR, a formal representation under a business
jurisdiction [11] is called a SBVR rule. SBVR rules can be of
two types: Strucutral Rule (used to expresses structure or
operation of a particular business entity) and Behavioural Rule
(used to express the conduct of a business entity).
III.
This section explains the used approach to automatically
map SBVR representation i.e. SBVR business rules to a UML
class model. To map SBVR to a UML class model, we have to
extract SBVR vocabulary from given SBVR rules and then
map the SBVR vocabulary to basic elements of a UML class
model (such as classes, associations, etc.) and finally generate
a graphical representation of class model. The used approach
works in following 5 phases:







SBVR2UML
MODEL TRANSFORMATION FROM SBVR TO UML




Figure 1. SBVR to UML Transformation Framework
All these five distinct phases are explained in detail in the
remaining part of the section.
A. Pre-Processing SBVR Specification
The input SBVR specification is pre-processed to find out
the candidates for possible SBVR vocabulary. We have user
the Stanford parts-of- speech (POS [17] tagger v3.0 to
tokenize and POS tag the input SBVR specification.
Then, the Stanford parser was used to syntactically analyze
the POS tagged SBVR specifications. The Stanford parser
generates the parse tree and typed dependencies. We have
used the extracted information by the Stanford parser for the
detailed semantic analysis. The semantic analysis of SBVR
specification is described in next section.
B. Semantic Analysis of SBVR Specifications
To identify the SBVR vocabulary, semantic role labeling is
performed. Semantic role labeling or thematic role labeling is
a common approach used in shallow semantic parsing. The
SBVR elements such as noun concept, individual concept,
object type, verb concepts, etc are identified from the SBVR
input. All these elements are shown in the extract of SBVR
(meaning) metamodel shown in figure 2:


Figure 2. A constrcut of SBVR (meaning) metamodel
Following set of mappings are used to extract the various
SBVR elements:
1) Extracting Object Types: In SBVR, the common nouns
(actors, co-actors, thematic objects, or beneficiaries) are
mapped as the object types [11] e.g. belt, user, cup, etc.
2) Extracting Individual Concepts: The proper nouns
(actors, co-actors, thematic objects, or beneficiaries) are
mapped to the individual concepts [11].
3) Extracting Fact Types: In SBVR specification, the
auxiliary and action verbs are represented as verb concepts. To
constructing a fact types, the combination of an object
type/individual concept + verb forms a unary fact type e.g.
“vision system senses”. Similarly, the combination of an
object type/individual concept + verb + object type forms a
binary fact type e.g. belt conveys part is a binary fact type.
4) Extracting Characteristics: In SBVR, the characteristic
[11] (section:11.1.2.2) or attributes are typically represented
SBVR
Metamodel
SBVR
Specification
UML
Metamodel
Class
Diagram
<<implements>>
<<ConformsTo>
<<ConformsTo>

Page 3

using is-property-of fact type e.g. “name is-property-of
customer”. Moreover, the use of possessed nouns (i.e. pre-
fixed by’s or post-fixed by of) e.g. student‟s age or age of
student is also characteristic.
5) Extracting Quantifications: In SBVR specification, all
indefinite articles (a and an), plural nouns (prefixed with s)
and cardinal numbers (2 or two) are mapped to the
quantifications.
6) Extracting Associative Fact Types: The associative fact
types [11] (section 11.1.5.1) are identified by associative or
pragmatic relations in English text. In English, the binary fact
types are typical examples of associative fact types e.g. “The
belt conveys the parts”. In this example, there is a binary
association in belt and parts concepts. This association is one-
to-many as „parts’ concept is plural. In conceptual modeling
of SBVR, associative fact types are mapped to associations.
7) Extracting Partitive Fact Type: The partitive fact types
[11] (section 11.1.5.1) are identified by extracting structures
such as “is-part-of”, “included-in” or “belong-to” e.g. “The
user puts two-kinds-of parts, dish and cup”. Here „parts‟ is
generalized form of „dish’ and „cup’. In conceptual modeling
of SBVR, categorization fact types are mapped to
aggregations.
8) Extracting Categorization
categorization fact types [11] (section 11.1.5.2) are identified
by extracting structures such as “is-category-of” or “is-type-
of”, “is-kind-of” e.g. “The user puts two-kinds-of parts, dish
and cup”. Here „parts‟ is generalized form of „dish’ and „cup’.
In conceptual modeling of SBVR, categorization fact types are
mapped to generalizations.
Fact Types: The
C. Mapping SBVR to Class Diagram
In this phase, overview of the transformations rules is
presented. These transformations rules are incorporated to
transform SBVR to UML class models. In Table I. the
informal correspondence in elements of SBVR and UML class
model is shown. Finally the SBVR rule is further processed to
extract the OO information.
TABLE I.
INFORMAL MAPPING BETWEEN SBVR AND UML METAMODEL
ELEMENTS
SBVR metamodel element
Object Type
Individual Concept
Characteristic
Verb Concept
Fact Type
Partitive Fact ct Type
Categorization Fact Type
Quantifications

The mapping of the each SBVR vocabulary item to the
respective UML class element is described below:
1) Mapping Object Type to Class: In a SBVR rule, all the
object types are mapped to classes in a UML class model.
UML metamodel element
Class
Object
Class Attribute
Class Method
Association
Generalization
Aggregation
Cardinalities
2) Mapping Individual Concepts to Objects: We are
mapping all individual concepts in a SBVR rule to the objects
in a UML class model.
3) Mapping Characteristics to Attributes: All the SBVR
characteristics or unary fact types (without action verbs)
associated to an object type are mapped to the attributes in a
UML class diagram.



Figure 3. A constrcut of UML (Class Diagram) metamodel
4) Mapping Verb Concepts to Methods: All the SBVR verb
concepts (action verbs) associated to a noun concept are
mapped to methods for a class e.
5) Mapping Associative Fact Types to Associations: A
unary fact type with action verb is mapped to a unary
relationship and all associative fact types are mapped to binary
relationships. The use of quantifications with the respective
noun concept is employed to identify multiplicity e.g. User(s)
and database will have one to many association in Figure 4.
The associated verb concept is used as caption of association.
6) Mapping Partitive Fact Types to Generalization: The
partitive fact types are specified as generalizations. The
subject-part of the fact type is considered the main class in
generalization and object-part of the fact types is considered as
the sub class.
7) Mapping Categorization fact Types to Aggregations: The
categorization fact types are mapped to aggregations. The
subject-part of the fact type is considered the main class in
aggregation and object-part of the fact types is considered as
the sub class.
D. Drawing UML Class Model
This phase draws a UML class model by combining class
diagram symbols with respect to the information extracted of
the previous phase. To draw a class diagram, three rectangles
were combined: one for class name, one for the class attributes
and one for the class methods. All the classes were
intelligently grouped. The classes having associations and
other relationships were drawn close to each other. Finally, the

Page 4

associations were also captioned with the titles and
cardinalities. The graphics functions in Java such as
(drawrect(), drawline(), etc) are used to draw the class diagram
and other symbols.
IV.
A CASE STUDY
An example of the software requirements for KeePass
Password Safe [18] is presented here. KeePass Password Safe
is an OSI Certified Open Source Software available under the
terms of the GNU license Ver. 2. Following is the problem
statement of the case study.

KeePass consists of a database which contains data for one
or more users. Each user’s data are divided into groups and
subgroups so that they are organized in a form that serves
right the user. Every user has a unique Master Key which can
be simple or composite and its combination opens uniquely
the database. If lost there is no recovery. Groups and
subgroups contain entries with usernames, passwords URLs
etc that can be sent or copied to websites, application and
accounts. There is also the ability for a onetime key creation
to be used once in a transaction without the risk of reused by
others for any reason.

We generated the SBVR rule representation of the above
input of KeePass Password specification. We have used
NL2SBVR tool [20] for generating SBVR representation. The
SBVR specification after extracting SBVR vocabulary is as
follows:

KeePass consists of a database which contains data for one or
more users. It is necessary that each user‟s data are divided
into groups and subgroups so that they are organized in a
form that serves right the user. It is obligatory that every user
has a unique Master Key which can be simple or composite
and its combination opens uniquely the database. If lost there
is no recovery. It is necessary that Groups and subgroups
contain entries with usernames, passwords, URLs etc that can
be sent or copied to websites, application and accounts. It is
possibility that there is also the ability for a onetime key
creation to be used once in a transaction without the risk of
reused by others for any reason.

Afterwards, the extracted SBVR vocabulary was mapped to
the UML class elements. Following information was extracted
in OO analysis phase:
TABLE II.
OBJECT ORIENTED ANALYSIS RESULTS
Example
Count
13
Details
Classes User, Database, Data, MasterKey, Simple,
Composite, Groups, SubGroups,
Applications, Webpage, Accounts,
OneTimeKey, Form
Attributes 03 user name, password, url
Methods 02 send(), copy()

Associations 06 database for user, form serves user, user
has Masterkey, its open database, Group
and Subgroup send or copy entries to
websites, applications, and accounts,
OneTimeKey used in a transaction

database contains data

data is divided into groups and subgroups,
MasterKey can be simple or composite

KeePass


Generalizations 01
Aggregations 02
Instances 01
There are two pieces of information such as “If lost there is
no recovery”, “without the risk” could not be processed as this
information should be translated to formal constraints such as
OCL. This could be achieved by using our other available tool
NL2OCLviaSBVR [19] available for free download. Another
issue was OneTimeKey should be sub-type of MasterKey but
this information has not been shown in generated UML class
model. Moreover, one construct “form serves the user” was
totally missed by the tool. By considering all these errors and
omissions the calculated recall was and calculated precision
was


Figure 4. A class model of case study generated by SBVR2UML
Figure 4 shows a screen shot of a class model generated
from extracted object oriented information of the input case
study “KeyPass Password Safe”.
V.
EVALUATION
To evaluate the performance of SBVR2UML tool, a set of
case studies including the case study discussed in section VI
were solved. The results of these case studies were used to
calculate recall and precision values as shown in table III.

Page 5

Average recall for SBVR requirement specifications to
UML class diagram transformation is calculated 83.82% while
average precision is calculated 91.017%. These results are
very encouraging for the future enhancements.
TABLE III.
SBVR2UML EVALUATION RESULTS
Example
Nsample
33
Ncorrect Nincorrect Nmissing
28 2
Rec% Prec%
Example 1 3 84.84 93.33
Example 2 40 37 2 1 92.50 94.87
Example 3 54 46 3 5 85.18 93.86
Example 4 38 31 3 4 81.57 91.17
Example 5 24 18 4 2 75.00 81.82
Average 83.82 91.01

We have also compared the results of SBVR2UML with
other available tools that can perform automated analysis of
the NL requirement specifications. Recall value was not
available for some of the tools. We have used the available
recall and precision values of the tools for comparison as
shown in table IV:
TABLE IV. A COMPARISON OF PERFORMANCE EVALUATION – SBVR2UML
VS OTHER TOOLS

NL Tools for Class Modelling
Recall Precision
CM-Builder (Harmain, 2003)
GOOAL (Perez-Gonzalez, 2002)
NL-OOML (Anandha, 2006)
LIDA (Overmyer, 2001)
UML-Generator (Bajwa, 2009)
SBVR2UML
73.00%
-
-
71.32%
-
83.82%
66.00%
78.00%
82.00%
63.17%
83.66%
91.01%

Here, we can note that the accuracy of other NL tools used
for information extraction and object oriented analysis is well
below than SBVR2UML. Moreover, the various tools‟
functionalities (if available, is automated or user involved) are
also compared with SBVR2UML as shown in Table IV:
TABLE V.
COMPARISON OF SBVR2UML WITH OTHER TOOLS
Support
CM-
Builder
LIDA GOOAL NL-
OOML
UML
2SBVR
Classes
Attributes
Methods
Associations
Multiplicity
Aggregation
Generalization
Instances
Yes
Yes
No
Yes
Yes
No
No
No
User
User
User
User Semi-NL
User
No
No
No
Yes
Yes
Yes
Yes
Yes
Yes
No
No
No
No
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
No
No
No

Table IV shows that besides SBVR2UML, there are very
few tools those can extract information such as multiplicity,
aggregations, generalizations, and instances from NL
requirement. Thus, the results of this initial performance
evaluation are very encouraging and support both the
approach adopted in this paper and the potential of this
technology in general.
VI.
CONCLUSIONS
In this paper, we addressed couple of challenging issues in
model transformation of SBVR specifications to UML class
models. First issue of SBVR parsing was addressed by using
typical NLP approaches. Second issue of transformation of
SBVR metamodel elements to UML class diagram metamodel
elements was addressed by using model transformation
technology. Sitra library was used for the purpose of the
model transformation. Moreover, the automated object
oriented analysis of SBVR specifications of software
requirements was performed. The results show that the
presented approach is a better approach as compared to the
other available approaches. Moreover, the SBVR2UML tool
provides a higher accuracy as compared to other available NL-
based tools. Besides better accuracy, SBVR has also enabled
to extract OO information such as association multiplicity,
aggregations, generalizations, and instances as other NL-based
tools can‟t process and extract this information.
In future, we aim to integrate our SBVR to OCL
transformation plugin [21] with SBVR to UML plugin, so that
a user may generate both UML and OCL with the same ease
and simplicity.
REFERENCES
[1] Bryant B.R, Lee, B.S., et al. 2008. From Natural Language
Requirements to Executable Models of Software Components. In
Workshop on S. E. for Embedded Systems:51-58.
[2] Ilieva, M.G., Ormandjieva, O. 2005. Automatic Transition of Natural
Language Software Requirements
Presentation. in proc. of Natural Language Processing and Information
Systems LNCS- 3513/2005:427-434mda/
[3] OMG. (2007). Unified Modelling Language (UML) Standard version
2.1.2. Object Management Group, Available at: http://www.omg.org/
[4] Overmyer, S.V., Rambow, O. 2001. Conceptual Modeling through
Linguistics Analysis Using LIDA. 23rd International Conference on
Software engineering, July 2001
[5] Perez-Gonzalez, H. G., Kalita, J.K. 2002. GOOAL: A Graphic Object
Oriented Analysis Laboratory. 17th annual ACM SIGPLAN conference
on Object-oriented programming, systems, languages, and applications
(OOPSLA '02), NY, USA: 38-39.
[6] Harmain, H. M., Gaizauskas R. 2003. CM-Builder: A Natural Language-
Based CASE Tool for Object- Oriented Analysis. Automated Software
Engineering. 10(2):157-181
[7] Oliveira, A., Seco N. and Gomes P. 2006. A CBR Approach to Text to
Class Diagram Translation. TCBR Workshop at the 8th European
Conference on Case-Based Reasoning, Turkey, September 2006.
[8] Anandha G.S., Uma G.V. 2006. Automatic Construction of Object
Oriented Design Models [UML Diagrams] from Natural Language
Requirements Specification. PRICAI 2006: Trends in Artificial
Intelligence, LNCS 4099/2006: 1155-1159
[9] Bajwa I.S., Samad A., Mumtaz S. 2009. Object Oriented Software
modeling Using NLP based Knowledge Extraction. European Journal of
Scientific Research, 35(01):22-33
[10] Ashfa Umber, Tayyiba Bashir, M Shahid Naweed, Imran Sarwar
Bajwa (2011) Requirements Elicitation Methods In:
International MIMT Conference (MIMT 2011) 541-545 Singapore
Specification into Formal
2nd IEEE

Page 6

[11] OMG. 2008. Semantics of Business vocabulary and Rules. (SBVR)
Standard v.1.0. Object
http://www.omg.org/spec/SBVR/1.0/
[12] Raj, A. , Parbharkar, T.V. Hendryx, S. 2008. Transformation of SBVR
Business Design to UML Models. India Software Engineering
Conference (ISEC 2008 ), ACM, Hyderabad, India
[13] Nemuraite, L., Ceopniene, l., et al. 2008. Representation of Business
Rules in UML&OCL models for developing Information Systems. First
IFIP Working Conference, PoEM 2008, Stockhom, Sweden.
[14] Kleppe, A., Warmer, J., Bast, W.: MDA Explained: The Model Driven
Architecture{Practice and Promise. The Addison-Wesley Object
Technology Se-ries. Addison-Wesley (2003)
[15] OMG: MOF Core v. 2.0
http://www.omg.org.
[16] Akehurst, D.H., Bordbar, B., Evans, M.J., Howells, W.G.J., McDonald-
Maier, K.D.: SiTra: Simple transformations in java. In Nierstrasz, O.,
Whittle, J., Harel, D., Reggio, G., eds.: Model Driven Engineering
Languages and Systems, 9th In-ternational Conference, MoDELS 2006.
Volume 4199 of LNCS., Genova, Italy,Springer (2006) 351{364
[17] Toutanova. K., Manning, C.D. 2000. Enriching the Knowledge Sources
Used in a Maximum Entropy Part-of-Speech Tagger. In Joint SIGDAT
Conference on Empirical Methods in Natural Language Processing and
Very Large Corpora: 63-70.
[18] Elia Kouzari, 2008. Software Requirements Specifications for KeePass
Password Safe. Software Engineering, Aristotle University Thessaloniki,
Available at: http://keepass.info/extensions/base/docs/Software
RequirementsSpecification-KeePass-1.10.pdf
[19] Imran S. Bajwa, Behzad Bordbar, Mark G. Lee [2010] "OCL
Constraints Generation from Natural Language Specification", in EDOC
2010 - IEEE International EDOC Conference 2010, Vitoria, Brazil, Oct
2010, pp:204-213
[20] Imran S. Bajwa, Mark G. Lee, Behzad Bordbar [2011] SBVR Business
Rules Generation from Natural Language Specification. in proceedings
of AAAI 2011 Spring Symposium -AI4BA, San Francisco, USA, Mar
2011, pp:2-8
[21] Imran S. Bajwa, Mark G. Lee [2011] "Transformation Rules for
Translating Business Rules to OCL Constraints", in ECMDA-FA 2011-
7th European Conference on Modelling Foundations and Applications,
Birmingham, UK, Jun 2011, pp:132-143
Management Group, Available:
Document Id: formal/06-01-01.