A Knowledge Base Guided Approach for Process Modeling in

Complex Business Domain

Roberto Paiano and Adriana Caione

Department of Engineering for Innovation, University of Salento, 73100 Lecce, Italy

Keywords: Knowledge Base, Business Process Management, Compliance, Natural Language Text.

Abstract: The business process analysis requires an in-depth knowledge of factors such as the activities carried out;

the actors involved; the domain or business context in which the activities are performed; the internal

company structure; the current regulatory framework. This involves the employment and the collaboration

of different professionals, such as business experts, domain experts and legal experts, along with a

considerable effort in terms of time and resources. For the purpose of an efficient and effective management

of business processes, it is also important to ensure the compliance with the company context and the

flexibility with regard to changes that may occur within the company or at the legislative level. This paper

shows a methodological and architectural approach guided by a knowledge base that describes the

application domain. It is populated iteratively with the information extracted from the analysis of

documents, regulations and requirements. The knowledge base is then used by the process designer as a

guide for business process modelling and management.

1 INTRODUCTION

Business process modelling and management is a

discipline that aims to increase the efficiency and

effectiveness of the activities carried out in the

companies. Indeed, it analyses the processes, which

is the company way of working, and identifies, by

comparison with the best practices in the same

context, the strengths and the weaknesses. It follows

a business process reorganization and reengineering.

The analysis of business processes requires an

in-depth knowledge of some aspects such as the

activities carried out in the company; the actors

involved; the domain or business context in which

the activities are performed; the internal company

structure; the current regulatory framework. This

implies the interaction and the collaboration of

different professionals, such as business experts,

domain experts and legal experts, along with a

considerable effort in terms of time, about 60% of

the total time spent to manage the processes (Herbst

and Karagiannis, 1999).

For the purpose of an efficient and effective

business process management, it is also important to

ensure compliance with the company context and

flexibility with regard to changes that may occur

within the company or at legislative level.

As reported verbatim in (Sadiq and Governatori,

2015) “The ever-increasing obligations of regulatory

compliance are presenting a new breed of challenges

for organizations across several industry sectors.

Aligning control objectives that stem from

regulations and legislation with business objectives

devised for improved business performance is a

foremost challenge”. Indeed, a prompt response to

changes and the compliance with the company

context could be the reason of improvement in the

competitive position compared to other companies

in the same sector.

In view of these consideration, there is the need

of methods and technologies able to facilitate and

speed up the acquisition phase of the information

related to the company context, identify and model

business processes and ensure the process

compliance with the company operational context,

even in the faces of changes.

Most of such business information, about 85%

(Blumberg and Atre, 2003), is stored in an

unstructured way, in text documents and the amount

of available unstructured sources is continuously

growing. But how these sources can be used for

defining and modelling business processes?

In this paper we propose a methodological and

architectural solution that falls in research

Paiano, R. and Caione, A.

A Knowledge Base Guided Approach for Process Modeling in Complex Business Domain.

DOI: 10.5220/0005974801690176

In Proceedings of the 11th International Joint Conference on Software Technologies (ICSOFT 2016) - Volume 1: ICSOFT-EA, pages 169-176

ISBN: 978-989-758-194-6

169

disciplines such as knowledge management

(Tiwana, 2000), process mining (Van Der Aalst,

2011) and business process management. It is based

on an approach guided by a knowledge base realized

by the help of domain and legal experts. This

knowledge base describes the entire company

domain in the form of concepts and relationships

among them. The idea is to populate, iteratively, the

knowledge base with the information extracted from

unstructured sources, and to use the knowledge base

for business process modelling, ensuring the

compliance with the company context and

facilitating the process adaptation to changes.

The purpose is to make the process design

experts as autonomous as possible during business

process management phases.

The paper is organized as follows. In Section 2

we describe the background of this research work.

Section 3 shows how the literature faces problems

similar to those analysed here. Section 4 describes

the methodology and the phases that characterize it.

Section 5 shows the approach from the architectural

point of view and of the components that

characterize it. Finally, Section 6 includes some

conclusions.

2 BACKGROUND

The contribution described in this paper represents a

continuation of the research work (Caione et al.,

2015a) which results in a methodological and

architectural solution able of managing the entire

business process life cycle in complex business

domain, from process modelling to process running

on an automatically generated web application

prototype.

First it is important to consider and know the

operational context in which the company takes

place in order to proceed with the definition of the

activities and business processes. To this end it is

essential a strong interaction between different

professionals. Because of a different cultural

background, they should use a common dictionary to

allow the knowledge sharing and transfer. In this

regard, in (Guido et al., 2015) the authors suggest

some guidelines addressed to legal experts in order

to schematize the texts of laws and regulations and

to simplify the understanding to the process design

experts who will model the processes in accordance

with the existing laws. The output of the guidelines

is a set of schemes that could provide support for

domain knowledge base modelling. The knowledge

base is actually the result of a careful analysis of the

operational context, the detection of the main

concepts that characterize it along with the

identification, reuse and merge of one or more

existing knowledge bases able to describe the entire

reference domain. In (Paiano et al., 2015) the

authors propose a knowledge model structured into

two ontological levels:

 Enterprise Domain Ontologies (EDO) that

“model the entities (i.e., human resources, tasks,

processes, objectives, etc.) that represent the

organizational environment from a general

perspective. Because of their general purpose,

these ontologies have been defined extending

and combining ontologies (i.e. reference

ontologies) already existing at the state of the

art”;

 Application Domain Ontologies (ADO) that

“model domain specific concepts and they

support the classification of entities represented

through the EDO. The regulations represent a

typical example of ADO because they are

application specific”.

This model is useful during the design phase of new

business processes compliant with the reference

domain and during the management of expected

and/or unexpected events.

With regard to the definition and management of

business processes, in the literature there are several

workflow management systems that support the

business and IT expert during the process planning,

execution and management. A survey is given in

(Caione et al., 2015b) in which the authors describe

five open source workflow management systems:

 Intalio BPMS (www.intalio.com);

 BonitaSoft (http://it.bonitasoft.com);

 jBPM (www.jbpm.org);

 Activiti BPMN 2.0 (www.activiti.org);

 Camunda (www.camunda.com)

These solutions are compared on the basis of eight

key aspects:

 process editor oriented to business and/or IT

experts;

 APIs availability for business and/or IT experts;

 connectors availability;

 BPMN 2.0 compliance;

 fast prototyping;

 prototyped application customization;

 process monitoring;

 engine source code customization.

Particular attention is paid to the requirements of

extending the solution in order to allow the

integration and the interaction with the domain

knowledge base.

ICSOFT-EA 2016 - 11th International Conference on Software Engineering and Applications

170

From the analysis and the obtained results, jBPM

has proved to be the most suitable system from the

point of view of the requirement fulfilment.

The processes are modelled using the standard

Business Process Model Notation (BPMN) also

available in version 2.0 (OMG, 2011). It includes

four categories of elements necessary for the design

of a process:

 flow Objects (Activities, Events and Gateways),

swimlanes (Pools and Lanes);

 artifacts (e.g. Data Objects, Text Annotations or

Groups);

 connecting objects (Sequence Flows, Message

Flows and Associations).

Typically, a process design reflects the analysis of

the operating context and of the company

requirement that is carried out through interviews,

meetings, etc., then transcribed into text form, along

with the understanding of structured and

unstructured documents.

The methodological and architectural approach

described in this paper, shows the use of these

guidelines, knowledge bases and tools. We have also

done a scouting on the state of art of software

solutions able to map the information contained in

unstructured documents with the concepts of a

previously modelled knowledge base, the tool

named OwlExporter was identified (Witte et al.,

2010).

It is an extension of the General Architecture for

Text Engineering (GATE) framework (Bontcheva et

al., 2004). It includes support for the semantic

language Ontology Web Language – Description

Logic (OWL-DL), the automatic population of a

knowledge base and the integration of information

taken from Natural Language Processing (NLP) in

order to make a more accurate reasoning.

Being the tool dependent on a specific

knowledge base, for the purposes of this work it will

have to be extended and customized for the specific

reference domain in order to better support the

process designers in the business process modelling

phase.

3 RELATED WORKS

Recently there is a considerable attention to the

identification and generation of process models from

content written in natural language.

An interesting solution is described in (Friedrich

et al., 2011). In the paper the authors present an

automatic approach for the generation of process

designs according to the BPMN standard and

starting from texts written in natural language.

Verbatim, they "combine an extensive set of tools

from natural language processing (NLP) in an

innovative way and augment it with an anaphora

resolution mechanism". The approach involves three

steps:

 Sentence Level Analysis, through which it is

possible to extract actors and actions (verb +

object) and connect the actors to the actions;

 Text Level Analysis, through which it is possible

to analyse the sentences taking into account their

mutual relationships. The end result is the

creation of flows of activities describing how the

activities interact with each other;

 Process Model Generation, the information

extracted in the previous steps and stored in a

template are transformed into processes

according to the BPMN standard.

Another solution, validated on a case study in the

archaeological field, is described in (Viorica Epure

et al., 2015). It is called TextProcessMiner and

consists of the generation and analysis of process

activity logs starting from text documents in an

attempt to define the process instance model. The

solution is "fully unsupervised and uses natural

language processing techniques with a focus on the

verb semantics" and is divided into three

components:

 TextCleaner, “responsible for cleaning the

methodology section and preparing the text

which will be mined”;

 ActivityMiner, “responsible for mining the

activities from the text”;

 ActivityRelationshipMiner, “responsible for

mining the relationships between the activities,

thus discovering the process instance”.

Lastly, we quote the paper (Brandão et al., 2015)

where the authors propose a solution able to solve

the problem of subjectivity related to the business

expert vision. It also facilitates the process

management, the reuse, the maintenance and the

understanding. The idea is to automatically identify

particular elements in business processes through the

analysis of the process event log. Within the

business process models, as defined by the authors,

"elements can be scattered (repeated) within

different processes, making it difficult to handle

changes, analyze process for improvements, or

check crosscutting impacts". These elements are

called aspects.

Some of the above mentioned solutions use

knowledge bases to define a set of useful rules in

order to analyse natural language texts. In this paper,

A Knowledge Base Guided Approach for Process Modeling in Complex Business Domain

171

the knowledge base plays a crucial role since it

describes the entire reference domain in terms of

concepts and relationships and it is integrated into

the business process management system to support

the business expert and the process designer during

the modelling, the reuse and the maintenance phases

of processes. Therefore, the option of automatically

populating the knowledge base from the information

contained in analysis and requirement documents

could support the business experts and process

designers, reducing their knowledge gap and the

time spent in the analysis of the operational context.

An attempt similar to that described in this paper,

though with different aims and tools, is shown in

(Rashwan et al., 2013) which illustrates an automatic

requirement classification system, based on support

vector machines. It automatically categorizes

requirement sentences into different Non-Functional

Requirements knowledge base concepts.

4 METHODOLOGY

The methodology adopted in this work consists of

three main phases that can be summarized as

follows:

 knowledge acquisition about the company

domain. The output of this phase is a knowledge

base that summarizes the context in the form of

concepts and semantic relationships among

them;

 information mapping, which aims to

automatically populate the knowledge base,

defined in the previous step, through the

information contained in the analysis, regulatory

and business requirement documents;

 business process modelling, the final result of

which is the BPMN representation of the

business processes modelled by the process

designer through the information contained in the

populated knowledge base.

In the following we describe in detail these phases.

4.1 Knowledge Acquisition about

Company Domain

In recent years, the use of knowledge bases has seen

a strong increase as a result of the spread of

semantic information systems. The design of these

systems is an activity demanding from the

standpoint of time and human resources. This led to

the definition of methodologies with the attempt to

standardize the construction of knowledge bases and

to achieve a result in a common semantic language,

in order to ensure the knowledge base sharing and

reuse even in different domains.

There are numerous methods in the literature for

the creation and the management of knowledge

bases, among them Methontology (Fernández-López

et al., 1997), Skeletal Methodology (Uschold and

King, 1995), Grüninger and Fox (1995), to name but

a few. They provide a valuable support to

knowledge engineers and carefully describe the

steps to be followed during the phases of modelling,

implementation and maintenance of the knowledge

bases.

Our approach inherits what realized in (Paiano et

al., 2015). The authors chose the methodology of

Grüninger and Fox, both because it is based on a

complete existing technique, Methontology, which

describes the entire knowledge base creation life

cycle, both because it extends this methodology and

focuses on the creation of knowledge bases from the

the information related to the application domain.

Added to this is the importance of the research

and the reuse of existing knowledge bases or parts of

them as additional support for the knowledge

engineer.

We combine the above mentioned work with

some helpful guidelines with the attempt of

schematizing complex domains such as those that

provide laws and regulations (Guido et al., 2015).

The end result is an abstract representation of the

application domain in a standard semantic language.

This representation organizes the company

knowledge in both ADO and EDO levels, briefly

described in the Background section.

Although there are tools that can semi-

automatically generate knowledge bases starting

from textual documents such as Ontogen (Fortuna et

al., 2007), Text2Onto (Philipp and Völker, 2005),

etc. For our purposes, they are not sufficiently

accurate to describe complex domains such as

business and regulatory ones. Therefore, from our

point of view, it is preferable that the construction of

such knowledge bases is guided by domain experts,

legal experts and business experts.

4.2 Mapping of the Information

Written in Natural Language

The implemented knowledge base, in order to be

helpful during the business process modelling phase,

should also contain the actual data extracted from

the analysis documents, regulatory and requirement

texts written in natural language.

Having built a solid knowledge base, it is

possible to use tools that, starting from documents

ICSOFT-EA 2016 - 11th International Conference on Software Engineering and Applications

172

written in natural language, instantiate the concepts

of the knowledge base. Among the tools we name

OwlExporter (Witte et al., 2010).

The main characteristics of this tool are:

 instance creation, using the entities contained in

the text;

 creation of the relations between concepts, using

the information and relationships in the text;

 coreference chains creation, leveraging the

concept of equality for the same entities that are

positioned in different parts of the text.

We have extended and customized this tool with the

idea of modelling business processes in complex

domains.

To this end, the pipeline and the source code

have been modified in order to overcome a

limitation which did not allow the processing of

complex texts.

Subsequently we have defined a lookup list of

entities and some JAPE transducers, that are files

containing some rules to further identify

patterns/entities from the text and to generate

annotations that will have to be matched with the

concepts and properties defined in the knowledge

base.

The benefit of automatically populating a

knowledge base from text, is to have a

categorization of the information contained in the

documents on the basis of the concepts modelled in

the knowledge base and, therefore, the possibility to

make queries and inference in order to extract

implicit knowledge.

Furthermore, taking advantage of the concepts

and instances contained in the knowledge base, it is

possible to enrich the model comparing the

information present in text documents with the

instances of the knowledge base.

4.3 Business Process Modelling

Since we have on the one hand a knowledge base

that schematises the application domain in terms of

concepts, instances and semantic relationships, and

on the other a business process management system,

the business expert and the process design are

facilitated in the design of the business activities.

More to the point, the interaction with the

knowledge base allows to fill the gap of the

knowledge expert in terms of the application

domain, speeds the information retrieval and

facilitates the adjustment of the processes if changes

happen in the application domain.

Our approach organises the business process

modelling into two main phases, as shown in Figure

1, Planning Time and Design Time.

In order to have a greater understanding of these

phases, it is good to define the concept of sub-

process template and the motivations of its use. It is

a block of BPMN elements, or a portion of process

that can not be executed by workflow engines, but

can be composed with other elements or portions to

derive executable business processes.

Figure 1: Business Process Modelling - Planning Time and

Design Time.

Planning Time - In this phase the process designer

uses an editor called Business Process Editor to

model the sub-process templates in BPMN standard

notation.

The editor allows the process designer to query

the domain knowledge base through an interface

named Knowledge Base (KB) Explorer. The

interaction is useful to semantically characterize the

sub-process template elements, establishing a

semantic relationship between the elements of the

sub-process templates and the knowledge base

concepts; to get some suggestions about the design

(the names of the process tasks, the input and output

parameters, etc.); to facilitate the management of

domain changes.

Design Time - In this phase the process designer

models the business processes that can be executed

within the workflow engines. To this end it may use

the previously defined templates, connecting them

according to three possible options:

 sequential adding: the connection in sequence of

the activities that constitute the sub-process

templates used;

 parallel adding: the connection in parallel of the

activities that constitute the sub-process

templates. This can be done using the parallel

A Knowledge Base Guided Approach for Process Modeling in Complex Business Domain

173

gateway of the BPMN notation;

 conditional adding: connection with branching of

the sub-process template activity flow. This is

applicable using all of the BPMN gateway types.

It is clear that the events and the redundant tasks

must be eliminated or merged.

At the end of the process composition and

modelling, the business designer must enter the

execution parameters, inputs and outputs to the

tasks, in order to make the processes executable. The

integration of such information may be done

querying the knowledge base.

The business model thus obtained is stored in

two different ways: in a format executable by

workflow engines (.bpmn extension); in a semantic

format so that it is sent to the knowledge base and

stored in it (.owl extension).

This makes easier to ensure the compliance of

the processes modelled with respect to the

application domain. A change in the domain, results

in a change in the knowledge base and, as a

consequence to a notification to the process designer

who will have to adapt the processes on which the

change has impact.

5 ARCHITECTURAL SOLUTION

The proposed architectural solution (Figure 2)

includes three main components.

Figure 2: Architectural solution.

 Knowledge Management System. It is responsible

for the management of the domain knowledge

base. It stores the concepts and the semantic

relationships and allows to make inferences and

queries. The access to the knowledge base is

realized, on the client side, using the interface

KB Explorer. The process designer can start a

search for concepts entering free text or

keywords. The interaction occurs through

Representational State Transfer (REST) and

Simple Object Access Protocol (SOAP) calls.

 Knowledge Base Population System. It is

responsible for the knowledge base population

with the information contained in unstructured

documents. It customizes and extends the

OwlExporter tool for the specific domain and is

based on GATE NLP framework that allows to

work with the knowledge bases using integrated

models.

The system is characterized by two knowledge

bases. The first models the concepts and the

relationships relevant for the particular domain.

The second knowledge base is a basic NLP

knowledge model independent from the domain

that contains the concepts commonly used in

language engineering, such as Document,

Sentence, Noun, Predicate, etc.

 Process Management System. It is responsible

for the business process modelling through the

composition of sub-process templates and it is

responsible for the entire management of the

modelled processes using the features offered by

the Knowledge Management System. The

Business Process System extends the jBPM

modelling editor for the features of sub-process

template definition, composition, saving of the

executable processes in a standard format and in

a format compatible with the knowledge base,

access to read and insert information to and in

the knowledge and management of domain

changes.

5.1 Process Management System

In this section we describe in more detail the main

modules that constitute the core component of the

architecture, the Process Management System.

The component conforms with the business 1

Knowledge Management System component

manages the domain knowledge base, result of the

knowledge acquisition about company domain phase

of the methodology. The Knowledge Base

Population System component allows the mapping

of the information written in natural language phase

of the methodology in order to automatically

populate the knowledge base with instances

extracted from documents.

Architectural details for the these two

components are provided in the papers (Paiano et al.,

2015) and (Witte et al., 2010), respectively.

ICSOFT-EA 2016 - 11th International Conference on Software Engineering and Applications

174

5.1.1 Knowledge Base Explorer

The KB Explorer module allows the interaction with

the knowledge base and the collection of the

concepts contained in it. The designer can retrieve

the concepts stored in the knowledge base by

performing one of the following search methods: by

free text; by keywords; by concepts related to the

results of a previous search.

This tool is integrated into the Business Process

Editor and can be used:

 to insert one or more references to the knowledge

base concepts, when creating a sub-process

template or a process;

 to characterize a specific BPMN element of the

model with one or more knowledge base

concepts, during the design of a process;

 to insert the input and output parameters

connected to the knowledge base concepts,

during the concretization of the process.

The concepts used in the models can assume

different states:

 active, when the concept is valid and can be used

to semantically describe an entire process or an

element of it;

 deleted, when the concept has been removed

from the knowledge base;

 replaced, when the concept has been modified in

the knowledge base because of some domain

changes.

The change in status is indicated by events from the

knowledge base.

5.1.2 Business Process Editor

The Business Process Editor is a standalone

application extending the Eclipse-based editor of

jBPM.

The necessity to update the source code is

dictated by the following features:

 creating a new model, that is a process or a sub-

process template. The wizard, then, allows the

process designer to use the KB Explorer in order

to select the concepts that will characterize the

design.

 opening an existing model to view it and/or to

make changes in it. The user can also update the

concepts of the model.

 modelling of a business process. In this activity

the process designer uses the tools, the graphic

elements, the palette and the properties of the

jBPM Eclipse-based editor. However, extensions

are needed to allow the process designer to

connect the process elements with the concepts

extracted from the knowledge base and to insert

the task execution parameters.

 saving the model. The model is saved in a format

executable by the jBPM workflow engine and it

can be exported in the corresponding semantic

representation. This semantic file is sent to the

knowledge base, transparently to the designer.

 management of changes in the application

domain. The process designer can check for

changes in the knowledge base, can see the

changed concepts and, then, he can update the

respective templates and/or processes.

 execution parameters entry. They are parameters

functional to execute a process. The definition of

the input and output parameters, through the

knowledge base, is obtained by invoking the KB

Explorer module.

 process recomposer. This is a useful feature for

events for the management of which is not

possible to identify a perfectly compatible

process. After the selection of an event, the

system suggests to the user a list of processes

and/or sub-process templates semantically

related to the event. The user can select one or

more processes and/or sub-process templates

among those proposed in order to insert and

connect them into the process that will handle the

event.

6 CONCLUSIONS AND FUTURE

WORKS

This paper presents a methodological and

architectural approach for business process

modelling, starting from the analysis and the

requirements of complex application domains.

Compared to the related works, where business

processes are obtained from documents written in

natural language, here a domain knowledge base

plays a central role in the business process

modelling phase. Indeed, it summarizes in terms of

concepts and relationships the application context

and it is integrated into a business process

management system. In this way the process

designer can proceed with the design of business

processes autonomously, without the need to interact

with domain and legal experts but with the

possibility to extract the required information from

the knowledge base.

Moreover, the use of an automatic knowledge

base populating system from text, constantly

A Knowledge Base Guided Approach for Process Modeling in Complex Business Domain

175

enriches this model and simplifies the reading and

the understanding of analysis and requirement

documents along with of more generally texts

describing the application domain.

Interesting is the possibility to define some

business-process templates that could be reused and

merged during the final process modelling phase.

These templates could be also semantically

characterized, tagging them and their elements with

the concepts of the knowledge base.

Finally, it is important to emphasize the

flexibility of the system with respect to the domain,

and consequently knowledge base, changes. The

proposed system, in effect, notifies to the process

designers the occurrence of such events and the

processes on which these events may have impact.

In the future it is expected to increase the level of

the system automation through the iterative use of

the system itself. We plan also to show some

examples and usage scenarios of the methodology

and the system on which to make evaluations and

comparisons of the obtained results with those

arising from the employ of a traditional approach,

that is without the use of a knowledge base, but with

the iterative interaction with the domain, legal and

business experts.

REFERENCES

Blumberg, R., Atre, S., 2003. The problem with

unstructured data. DM REVIEW, 13(42-49), 62.

Bontcheva, K., Tablan, V., Maynard, D., Cunningham, H.,

2004. Evolving GATE to meet new challenges in

language engineering. Natural Language Engineering,

10(3-4), 349-373.

Brandão, B. C. P., Santoro, F., Azevedo, L. G., 2015.

Towards Aspects Identification in Business Process

Through Process Mining. In Proceedings of the

annual conference on Brazilian Symposium on

Information Systems: Information Systems: A

Computer Socio-Technical Perspective-Volume 1 (p.

99). Brazilian Computer Society.

Caione, A., Guido, A. L., Martella, A., Paiano, R.,

Pandurino, A., 2015a. Knowledge base support for

dynamic information system management.

Information Systems and e-Business Management, 1-

44.

Caione, A., Guido, A.L., Paiano, R., Pandurino, A., 2105b.

A Survey of Open Source Workflow Management

System. International Journal of Emerging Trends &

Technology in Computer Science, ISSN: 2278-6856

VOL. 4, pp.22-26.

Fernández-López, M., Gómez-Pérez, A., Juristo, N., 1997.

Methontology: from ontological art towards

ontological engineering.

Fortuna, B., Grobelnik, M., Mladenic, D., 2007. Ontogen:

Semi-automatic ontology editor (pp. 309-318).

Springer Berlin Heidelberg.

Friedrich, F., Mendling, J., Puhlmann, F., 2011. Process

model generation from natural language text. In

Advanced Information Systems Engineering (pp. 482-

496). Springer Berlin Heidelberg.

Guido, A. L., Paiano, R., Pandurino, A., 2015. From laws

to business process: reducing the skill gap between

legal professional and business process analyst. In

Internet Technologies and Applications (ITA), 2015

(pp. 23-28). IEEE.

Grüninger, M., Fox, M. S., 1995. Methodology for the

Design and Evaluation of Ontologies.

Herbst, J., Karagiannis, D., 1999. An inductive approach

to the acquisition and adaptation of workflow models.

In Proceedings of the IJCAI (Vol. 99, pp. 52-57).

OMG, O., 2011. Business Process Model and Notation

(BPMN) Version 2.0. Object Management Group.

Paiano, R., Pandurino, A., Guido, A.L., Ritrovato, P.,

D’Apice, C., Laria, G., 2015. An Approach to

Integrated Management System Exploiting

Knowledge Base to Support Business Processes

Management. In Italian chapter of association for

information systems (ItAIS).

Philipp, C., Völker, J., 2005. Text2Onto-A Framework for

Ontology Learning and Data-driven Change

Discovery. In Proceedings of the 10th International

Conference on Applications of Natural Language to

Information Systems-NLDB (Vol. 5, pp. 15-17).

Rashwan, A., Ormandjieva, O., Witte R., 2013. Ontology-

Based Classification of Non-Functional Require-

ments in Software Specifications: A new Corpus and

SVM- Based Classifier. In The 37th Annual

International Computer Software & Applications

Conference (COMPSAC 2013), (pp. 381–386). IEEE.

Sadiq, S., Governatori, G., 2015. Managing regulatory

compliance in business processes. In Handbook on

Business Process Management 2 (pp. 265-288).

Springer Berlin Heidelberg.

Tiwana, A., 2000. The knowledge management toolkit:

practical techniques for building a knowledge

management system. Prentice Hall PTR.

Uschold, M., King, M., 1995. Towards a methodology for

building ontologies (pp. 15-30). Edinburgh: Artificial

Intelligence Applications Institute, University of

Edinburgh.

Van Der Aalst, W., 2011. Process mining: discovery,

conformance and enhancement of business processes.

Springer Science & Business Media.

Viorica Epure, E., Martin-Rodilla, P., Hug, C., Deneckere,

R., Salinesi, C., 2015. Automatic process model

discovery from textual methodologies. In Research

Challenges in Information Science (RCIS), 2015 IEEE

9th International Conference on (pp. 19-30). IEEE.

Witte, R., Khamis, N., Rilling, J., 2010. Flexible Ontology

Population from Text: The OwlExporter. In LREC

(Vol. 2010, pp. 3845-3850).

ICSOFT-EA 2016 - 11th International Conference on Software Engineering and Applications

176