USER AUTONOMY IN REQUIREMENTS CHANGING

SUPPORTED BY ORGANIZATIONAL SEMIOTICS AND

TAILORING

Carlos Alberto Cocozza Simoni, Maria Cecilia Calani Baranauskas

Institute of Computing, State University of Campinas, Av. Albert Einstein 1251, Campinas, Brazil

Rodrigo Bonacin

CenPRA - Rodovia Dom Pedro I, km 143,6, Campinas, Brazil

Keywords: Organizational Semiotics, Tailoring, Software Engineering.

Abstract: Nowadays, organizations face the impact of changes from several sources. Literature points out that we still

have a gap between the dynamic of the system maintenance and changes in the organizational processes. To

cover this gap we consider the use of practices coming from Organizational Semiotics (OS) and Tailoring.

OS allows a deep understanding of the organizational context and the technical system embedded in it, and

Tailoring proposes to provide autonomy to users in dealing with changes in computer systems. This paper

presents a case study developed in our University to explore and extend an existing approach to provide

more autonomy to computer system users, according to the evolution of their business requirements.

1 INTRODUCTION

Although more investment has been done in IT the

evolution of the existing computational systems

don’t follow the evolution of the organizational

requirements (Liu et al, 2002). Flexible architecture

in computational systems (Truex et al. 1999),

evolutionary Information Systems (IS) and co-

evolution between software and business process

(Liu et al. 2002) are efforts to cover this gap.

Organizational Semiotics (OS) allows us to

understand the role played by the IT in the

organizational contexts, and clarifies how an

organization interacts with the environment.

Tailoring (Morch 1995) proposes autonomy to users,

modifying the software according to the continuous

changes in business requirements.

In this paper we present a case study conducted

in our University to provide more autonomy to end

users in changing their applications according to the

evolution of their business requirements. The paper

is organized as follows: Section 2 presents some key

concepts of the OS methods and Tailoring, Section 3

presents the proposed approach, the case study and

discusses benefits and drawbacks of the proposed

approach, and Section 4 concludes the paper.

2 THEORETICAL

BACKGROUND

In this session we present the main concepts

involving Organizational Semiotics and Tailoring

applied to our work.

Organizational Semiotics (OS)

Social, cultural and organizational aspects involved

in the problem must have a more decisive role in the

process of developing the information system (Liu

2000). The MEASUR - Methods for Eliciting,

Analyzing and Specifying User’s Requirements

(Stamper, 1993) is a set of methods, based on OS, to

deal with the use of signs, their function in

communicating meanings (semantic), intentions

(pragmatic), and their social consequences (social).

In this work we use the Semantic and Norm

Analysis methods.

The Semantic Analysis Method (SAM) assists

analysts and users in eliciting and representing their

493

Alberto Cocozza Simoni C., Cecilia Calani Baranauskas M. and Bonacin R. (2006).

USER AUTONOMY IN REQUIREMENTS CHANGING SUPPORTED BY ORGANIZATIONAL SEMIOTICS AND TAILORING.

In Proceedings of the Eighth International Conference on Enterprise Information Systems - ISAS, pages 493-496

DOI: 10.5220/0002492004930496

 SciTePress

requirements in a formal and precise model. It

describes a view of agents with responsibilities in

the domain and their behavior patterns called

affordances.

The Norm Analysis Method (NAM) focuses on

social, cultural and organizational norms that govern

the actions of agents. A Norm defines a

responsibility of an agent engaged in a task, or

condition under which certain actions may (deontic

operators) be performed by the agent.

Tailoring

Morch and Mehandjiev (2000) consider that end

users of software should be responsible for the

process of changing a computer-based tool in its

usage context.

Morch (1995) classifies Tailoring in three

categories: Customization, Integration, and

Extension. The focus of our work is in Extension

where the functionality of an application is enhanced

by adding new code. This author considers that

Extension is a not common feature of commercial

applications, and the main question pointed out by

him concerns the implementation language to be

used in this scenario, which should be accessible to

end users.

He also considers that Tailoring techniques could

be extended to support organizational changes. We

argue that OS may provide methods suitable for

doing Tailoring, since it puts equal emphasis on both

technical and organizational issues.

3 AN APPROACH PROVIDING

AUTONOMY

In this session we present an approach to provide

software autonomy to specialists and end users in

changing their applications according to their needs.

Based on previous works (Bonacin and

Baranauskas 2005; Simoni et al. 2005 a/b) we use

OS to provide a contextual understanding of the

problem, to model the static characteristics of the

context (SAM), and its dynamic aspects (NAM).

Based on Tailoring and OS concepts, a norm-

based framework called NBIC – Norm-Based

Interface Configurator was structured (Bonacin and

Baranauskas 2005) to support the work of interface

specialists. In the following sessions we present this

framework and the case study carried out in our

University.

3.1 NBIC – Norm-Based Interface

Configurator

By using NBIC architecture, the system maintenance

is understood in tree levels (Figure 1):

1. End users (or domain specialists) use tailorable

interfaces to change the system behavior;

2. An interface engineer interacts with three tools:

(a) Norm Manager to see the results of the

tailoring activity at level 1, (b) Action Manager

to link norms actions to interface objects and

actions, and (c) Tailoring Manager to specify the

tailoring interface to be used by the end-users.

3. The code of the system static part could be

maintained.

The users’ participation is not restricted to a

specific level. They should cooperate with the

interface engineer in the second level and with the

developers in the third level.

The end-user applications interact with the NBIC

by two mechanisms: perception and action. The

“perception” mechanism is the part of the system

that transmits context information to the inference

machine using a specific protocol. The “action”

mechanism reads the parameters resulted from the

inference machine that evaluate the specified norms

and triggers the appropriated action.

3.2 Extending NBIC with Norm

Modeler: a Case Study

We have structured a course (Special Topics in

Information Systems) in our University allowing the

students to have contact with the theme of the

evolving computational systems. The main goal was

to provide an overview about ways to deal with the

existing dependencies between the software and

organizational changes. The focus of our case study

was the first level of NBIC showed in Figure 1,

where the end-user can directly change the

application.

The class was divided in 9 teams with 4 students

in each one. The students’ groups worked with the

conception and the construction of two prototypes:

(1) a business application to be tailored and (2) a

norm modeler.

The method allowed the students to model the

problem context in terms of static (SAM) and

dynamic elements (NAM). All teams succeeded in

modeling the context through agents, their behavior

and norms that guide them.

Our interest was in investigating what kind of

software flexibility we could get and the end-user

skill that would be necessary to use this scenario.

ICEIS 2006 - INFORMATION SYSTEMS ANALYSIS AND SPECIFICATION

494

Each team chose its own technical platform,

allowing us to verify the independence between the

business solution and the Norm Modeler.

The norm modeler was used as the tailoring

interface box represented in Figure 1. The tailoring

interface is a framework that can be configured by

the tailoring manager tool, in order to provide end-

users ways of specifying norms using direct

manipulation. The Norm Modeler was not restricted

to this framework; the students could explore any

technique to provide an interface where end-users

could specify norms. The main focus was on the

construction of applications that are easy to maintain

by using the tailoring and the OS theoretical

referential.

To allow the interaction between the business

application and the Norm Modeler the students

realize the necessity of a common dialog between

them. They proposed a set of interface elements that

should be provided and managed, such as: agents,

affordances, attributes, rule names, actions, interface

objects. To allow non-specialist users in computing

to interact in this scenario, the students considered

that, for each element, it would be necessary to

allow users to assign names, descriptions and

graphical representation, registering their own

expression of each element.

3.3 The Norm Modeler

The main interface proposed by Team 1, which had

achieved the most complete solution for the Norm

Modeler, is shown in Figure 2. The features

developed are marked with capital letters:

A. To chose a norm name.

B. To specify the conditions to activate the norm.

C. To specify the deontic effect of the norm,

permitting/obligating/prohibiting agent actions.

D. To specify the agents involved in the action.

E. To specify the agent actions that should be

performed, according to the deontic effect.

F. To show to the end-user the result of the norm

specification.

Each feature described above is activated and

controlled by through the software that supports the

interface. In Figure 2 we can also see an interface to

allow the end-user to describe each element involved

in the norm specification, to associate some

graphical representation, and to write expressions.

3.4 Validation and Discussion

To evaluate the solutions proposed, each team that

designed a Norm Modeler tested their solution with

one of the norm-based applications designed to

support the chosen business problem. We could

verify that the business interfaces and the norm

modelers designed had a good independence

between them, in terms of technology employed and

design process.

The majority of the students considered that a

norm-based modeling, as provided by Semantic and

Norm Analysis from OS, was an important factor to

provide this independence and autonomy.

The students with some experience in working

with business organizations, had considered that the

software designed could be used by specialists in the

business domain, with some experience with

process, notations etc.

Figure 1: NBIC Architecture Overview.

USER AUTONOMY IN REQUIREMENTS CHANGING SUPPORTED BY ORGANIZATIONAL SEMIOTICS AND

TAILORING

495

Figure 2: The main Norm Modeler interface from Team 1, and an example of interface to facilitate element description.

4 CONCLUSION

With this work we have verified the possibility to

support users in dealing, by themselves, with

changes in software resulted from modifications in

the business process and their requirements.

Integrating OS and Tailoring concepts and

techniques allowed us to think about a development

environment where business and technical rules,

constraints, and actions could be modeled and

described in terms of the semantic employed by the

agents involved in the organizational context.

The final results encourage further work towards

exploring this approach in practical work, using

actual business scenarios, allowing us to verify the

desired user autonomy and the influence of the

approach in the quality of the software application,

and in the business process as well.

REFERENCES

Bonacin, R. and Baranauskas, M. C. C., 2005. An

Organizational Semiotics Approach Towards

Tailorable Interfaces. Proceedings of 11th

International Conference on Human-Computer

Interaction, 2005, Las Vegas.

Morch, A., 1995. Three Levels of End-User Tailoring:

Customization, Integration, and Extension. Third

Decennial Aarthus Conference, August 14-18 Aarthus,

Denmark, p.41-51.

Morch A. and Mehandjiev N., 2000. Tailoring as

Collaboration: the Mediating Role of Multiple

Representations and Application Units. Kahler, H,

Morch, A., Stiemerling, O., Wulf, V. (Eds), Special

Issue on Tailorable Systems and Cooperative Work,

CSCW, Kluwer, v. 9, Issue 1.

Liu, K., 2000. Semiotics in Information Systems

Engineering, Cambridge: Cambridge University Press.

Liu, K., Sun, L. and Bennett, K., 2002. Co-Design of

Business and IT Systems – Introduction by Editors.

Information Systems Frontiers, 4(3), p. 251-256.

Simoni, C.A.C., Bonacin, R. and Baranauskas, M.C.C.,

2005a. Bringing Social Constructs to the Information

System Development Process: Contributions of

Organizational Semiotics. In 7th ICEIS - International

Conference on Enterprise Information Systems. 25-28

May. Miami, EUA, p.112-119.

Simoni, C.A.C., Melo A. M. and Baranauskas, M.C.C.,

2005b. Towards a Social-based Process for

Information System Development: A Case Study. In

The 8th International Workshop on Organizational

Semiotics. 23-24 June. Tolouse, França, p.135-149.

Samper, R.K., 1993. Social Norms in requirements

analysis – an outline of MEASUR. In: Jirotka, M.,

Goguen, J. and Bickerton, M. (eds.), Requirements

Engineering, Technical and Social Aspects. New

York: Academic Press, apud LIU, K., 2000. Semiotics

in Information Systems Engineering, Cambridge:

Cambridge University Press.

Truex, D.P., Baskerville, R. e Klein, H., 1999. Growing

Systems in Emergent Organizations: As the new

economic realities pressure your organization to

change from stable to emergent, new practices for IT

support are required. Communications of the ACM,

August 1999, 42 (8), p.117-123.

ICEIS 2006 - INFORMATION SYSTEMS ANALYSIS AND SPECIFICATION

496