Intelligent Agents and Apache Cocoon for a CV Generation System

Evelio J. Gonzailez, Alberto Hamilton, Lorenzo Moreno, Juan A. Mendez, G.N. Marichal, Jose Sigut,
Marta Sigut and J. Felipe
Dpto. Ingenieria de Sist. y Automatica y Arq. y Tecnologia de Computadores
Universidad de La Laguna, CP 38207, La Laguna, SPAIN
E-mail: ejgonzal@ ull. es

Abstract forms are generated in a dynamic way: the structure of the

form should change as soon as the data architecture is
A dynamic system - based on Multiagent Systems modified, as several fields can be added in the CV data
(MAS) with XML and Apache Cocoon - for the automatic request by the official institutions. In addition to this, the
and dynamic generation of CV documents is presented in data introduced through those forms should be validated
this paper. This pretends to be a useful application in against the restrictions codified in XML.
particular for an academic and research environment, This system for the dynamic generation of CV
since every University model usually requires its members documents has been integrated in a Multiagent System,
to manage a big amount of personal data, such as originally developed for the automatic management of
publications in journals and conferences attended. The agendas in a Universitary Scenario [1][2][3]. This system
application presented in this paper appears as a web has been designed using FIPA specifications [4] and its
portal where the users - in this case, professors and aim is to help the members of the Universitary Scenario to
research students- can manage their CV data, where the organize internal meetings and to get resources such as
authors remark the dynamic generation of the web portable computers and projectors. For the integration of
environment. the web environment into that system, the authors have
implemented several agents whose behaviour can be
1. Introduction studied in an independent way. In addition to this, the
agents in MASplan, and therefore the new agents, take the
The Spanish university model requires its members to advantage of the use of ontologies, expressed in a highly
manage a big amount of personal data, such as expressive language, OWL.
publications in journals and conferences attended. The reason for the use of these technologies lies in the
Different official institutions often require researchers and human behaviour. Firstly, it has been observed that
students for these data in order to different purposes such sometimes the corresponding author of an article forgets
as awarding a contract or research fellowship, etc. to communicate the acceptance of that paper to the rest of
Unfortunately, it is usual that each institution has its own the authors. Even if the corresponding author sends, e.g.,
template to fill, so researchers are often condemned to an e-mail to a co-author communicating the good news,
waste their time typing the same data in different this co-author is usually so busy that he/she prefers to
documents. It has been calculated that the generation of a update his/her CV database later, taking the risk of
CV takes an average 3-hour period and that an automated 'losing' the paper in his/her CV. This behaviour implies
system could save at least $25,000 per 100 generated that each user sends an e-mail to every related colleague,
CV's. looking for that 'lost' publications, whenever he/she needs
This scenario immediately brings XML into one's to present a CV document, wasting a lot of time in this
mind. Users would only need to type their data once in a way. Thus, a general database - an initial attempt was
XML file and apply a different XSL transformation in initially implemented in Apache Cocoon-, covering all the
order to generate each type of document. Unfortunately, users' items for CV, could be a good solution. However,
there are a good number of users in this scenario that find once more the authors have bumped into the human
XML extremely difficult. Thus, it would be desirable to behaviour. In spite of the security offered by the database
have a user-friendly web environment in which manager, a significant number of members of the
researchers and students could manage their personal data Universitary Scenario were reluctant to insert their data,
and generate their updated CV with only a click. claiming that they did not want leave their data in a
An interesting tool for that purpose is Apache Cocoon, centralized system where they could be accessed by
and in particular its forms (Cocoon forms or CForms), a malicious people. In this context, the features provided by
XML way to build forms that can be filled by the users. the multiagent systems - distribution, reliability,
One of the main contributions of this paper, is that the

1-4244-1031-2/07/$25.00©2007 IEEE 9
proactivity, autonomous and reactive behaviour, etc.- communication because a shared ontology provides the
seem to be especially useful. common format in which express data and knowledge. An
ontology is usually defined as a set of classes, relations,
2. State of the art in MAS, Apache Cocoon functions, etc. that represents knowledge of a particular
domain that provides the common format in which express
and Generation of CV documents. data and knowledge [6]. There are several ontology
languages. However, from their experience, the authors
This section is dedicated to describe the state in the art of prefer those languages known as "markup languages",
the main elements of the development of the system. whose last generation (RDF, DAML+OIL [7] and OWL)
2.1 Multiagent systems and ontologies gives computers an extra small degree of autonomy that
can help them to do more useful work for people. Systems
Agents and Multi-Agent Systems [1,2] (MAS) are part may be able to provide all sorts of additional services and
of a new programming paradigm. They have been responses beyond the requirements of the standard but a
successfully used in a wide range of applications such as certain basic set of conclusions will always be required.
robotics, e-commerce, agent-assisted user training, In this sense, OWL is on the top of the ontologies
military transport or health-care. However, why using languages, although there are other markup languages,
agents? This is not a trivial question, as there is not a such as DAML+OIL that are sufficiently expressive for
uniform and widely accepted definition of agent3. carrying out the project described in this paper.
Examples of agent definitions are: Unfortunately, OWL does not allow developers to
Autonomous agents are computational systems that express all the constraints they would like, for instance,
inhabit some complex dynamic environment, sense and those related to the limitation in the range of a value or
act autonomously in this environment, and by doing so constraints between values. In this way, A MAS system
realize a set of goals or tasks for which they are designed. could manage these situations, for instance, taking
(Maes) advantage of the fact that OWL datatype properties may
Intelligent agents are software entities that carry out make use of simple types defined in accordance with
some set of operations on behalf of a user or another XML Schema datatypes.
program with some degree of independence or autonomy, It is noted that the main aim of this paper is not to
and in so doing, employ some knowledge or justify the use of MAS and ontologies in a strict way
representation of the user's goals or desires. (IBM,) (there is a lot of literature about that fact), but to present
Autonomous agents are systems capable of the software experience when using them in the
autonomous, purposeful action in the real world. development of an example application. [5,6]
(Brustoloni)
As can be seen, agents are often characterized by 2.2 Apache Cocoon
describing their features (long-lived, autonomous,
reactive, proactive, collaborative, ability to perform in a This is a mature and popular web development
dynamic and unpredictable environment). Users can framework, developed in Java and based on the Servlet
delegate to agents tasks that are designed to be carried API model, that has been built around Separation of
without human beings intervention, for instance, as Concerns (SoC) and component-based development
resource managers or personal assistants that learn from (COP), providing pipelined SAX processing.
its user. Nevertheless, in most of applications, a It introduces design patterns, evolutionary guidelines
standalone agent is not sufficient for carrying out the task: and software that make easy the creation of web services,
agents are forced to interact with other agents, forming a using a cache system in order to get a better performance.
multi-agent system (MAS). Due to their capacity of This technology, interfacing with many different back-end
flexible autonomous action, MAS can treat with highly and data formats, is focused on human resource
dynamic or uncertain environments. On the other hand, management rather then technological details, left in the
MAS can effectively manage situations where distributed middleware level. Its modular structure and Model-view-
systems are needed: the problem being solved is itself controller (MVC) architecture, based on XML files,
distributed, data are geographically distributed, managed makes the development extremely customizable with
by different control systems and/or difficult to share, minimal coding. Furthermore, there is a really plethora of
systems with many components and huge content. In this tools and XML derivates that can interact with Apache
sense, it is not exaggerated to say that MAS have become Cocoon.
a key tool for Al researchers and developers. From a practical point of view, Apache Cocoon uses a
In this context, having a shared ontology for MAS centralized configuration mechanism called the sitemap, a
inter-agent communication is critical to successful declarative XML document describing a set of pipelines
that will be invoked upon a URI pattern match. A pipeline

10
consists of three main components: a generator (which * The agents form a distributed system and it is not
produces SAX events), one or more transformers (which necessary a permanent connection. The agents are
operate on the SAX event stream, transforming it into who interact, not the users.
some other grammar) and a serializer (which transforms * Extensibility. Using both MAS and ontologies,
the SAX event stream into an output stream for the client new types of agents (or new instances of the same
browser or file) [9]. agents, even implemented by different developers)
In spite of Apache Cocoon started as a simple servlet can be added easily to the system, making its
for static XSL styling, it has become a powerful tool as functionality grow in a dynamic way. In general,
many features have been included. One of these features is this easiness cannot be reached by centralized
its CForms, an advanced forms framework that provides a systems, for example, a central server that every
solid basis for building interactive web applications. user interacts with via their Web browser.
Describing the structure of the forms involves the
definition of the widgets it consists of. A widget is an For the development, they authors adopted FIPA
object that knows how to read its state from a Request specifications because they have become a stronger
object, how to validate itself, and can generate an XML standard in MAS development and involves not only
representation of itself. Using widgets, developers can agent language specifications but agent management,
indicate, for example, that a specific field should contain conversations ...
an integer or a date [10]. In order to create a CForm, a FIPA (Foundation for Intelligent Physical Agents) is an
developer needs to define two XML files: a form model - organisation whose purpose is to promote the
that describes the structure of the form- and a form development of specifications of generic agent
template - that informs the place a widget is desired to specifications. Its agent management reference model
appear in the form. An optional third file is the form provides the normative framework within which FIPA
binding, which avoids having to write actual code for the agents exist and operate. The Directory Facilitator (DF)
edition of things like the properties of a bean or data from provides yellow pages services to agents that query it to
an XML document.[7,8,9] find out services offered by other agents. The Agent
Management System (AMS) offers white pages services
2.3 Generation of CVs and maintain a directory, which contain transport
addresses for agents registered in the Agent Platform
The problem of the automatic generation of CVs is not (AP). The Message Transport Service (MTS) is the
new. A good number of institutions and companies have default communication between agents on different APs
realized that users often waste a lot of time when writing, (FIPA Agent Management Spec.).
updating and - not less important- formatting their CVs. These specifications allow users not to be worried
For example, collecting the CVs of a relatively high about technical aspects such as a detailed communication
number of members of an institution for an official inform implementation. As indicated above, agent-based
and giving the impression of style homogeneity, could computing provides the decomposition, organization and
become in a heroic effort, even when strict guidelines abstraction of multifaceted applications in heterogeneous
were given. Thus, several developers have tried to solve networks.
this problem through different software. The analysed The authors have implemented a MAS for planning
attempts seem to be limited to an only style of CV and and scheduling in a University Research Group. This
they do not take any of the advantages of using the XML MAS, called MASplan, should help group members to
technology. Moreover, two of them are based on find the best possible time frames to perform a meeting
commercial products, so a user is laid when a structural and to designate the use of the common resources.
change in the application is desired. In this aspect the Originally the system MASplan was composed of 6
design of a system, based on XML and open-source different types of agents. The agents for CV generation
technologies is justified. have been integrated in this system, thus the authors
consider its brief description as illustrative.
3. MAS Architecture and Development User Agent (UA): This agent is an end-user interface,
which shows the schedule to its related user and allows it
The application of MAS to this problem is justified by to ask for a meeting or a resource. When it occurs, this
the following reasons. agent tries to locate its negotiator agent and communicates
* The environment is dynamic. For instance, the what user needs. Once the negotiator has finished its
number of users and their preferences can change work, the user agent receives the result and shows it to the
in an unpredictable way. The agents should adapt user.
themselves to these situations. Negotiator Agent (NA): The implementation of the
meeting and resource negotiation algorithm is applied via

11
this agent. When it is asked by its related user agent for a
meeting negotiation, it looks in the DF for the negotiator
agents of the rest of the intended attendees. Then, the
E;! CVSA OA1
negotiation process begins. Alternatively, in the case of a
resource negotiation, it looks for the resource agent.
querref (area)
Ontology Agent (OA): It provides ontology services to agree
an agent community, so that the identification of a shared
ontology for communication between two agents is
facilitated. The definition of an external ontology,
inform (area) query-if I
managed by an OA, provides numerous general
advantages: it permits consultation with regard to agree
concepts, the updating and use of ontologies and it l ~~ ~~<l
eliminates the need to program the entire ontology in I inform
every agent, hence reducing required resources.
Resource Agent (RA): This agent is invoked when a
queryref(CV)
resource negotiation occurs. Firstly, it asks the Ontology agree
Agent for the instances of the selected resource type.
Mail Agent (MA): When an agenda change is inform (CV) >1
confirmed, the Mail Agent is requested by the respective >1
negotiator agent to send an email to the user via the mail
software. For this purpose, it asks to OA for the email Figure 1: Message flow in the system
address of the user, as these data are stored in the
ontology. Universitary Scenario were reluctant to insert their data,
Rule Agent (RuA): This agent provides the system with claiming that they did not want leave their data in a
the ability of learning from the users. The RuA is centralized system where they could be accessed by
consulted whenever there is an agenda change in order to malicious people. Thus, each user owes its own database,
organize a meeting. The NA will consult with this agent stored in the local system and accessed through the
in order to determine whether or not the user is supposed Cocoon web portal environment with an authorization and
to agree to a possible agenda change. verification process. That database should be managed by
an agent in the system. In order to reuse code and
In this scenario, several mobile agents have been resources, the authors have implemented that management
integrated. Each user owes its own mobile agent, called functions in the UA code. Thus, the CVSA interacts with
CVSearchAgent (CVSA) that is periodically migrating in the corresponding UA when it is necessary. One of the
the network looking for new items for CV in which its actions to be carried out is to avoid redundancy in the
user is involved. It is clear from the nature of the scenario items for CV.
that the CVSA's do not need to be always active as users CVSA's have been implemented using JADE tool.
do not need to be continuously submitting their CV. Thus, This is a well-supported agent framework and the authors
their activity is reduced to a few hours each 7-15 days. consider that its use is more adequate for this
This fact and the characteristic of mobility make that the implementation, as it allows a better management of the
network is not overloaded, as the interaction with other agent life cycle (active/inactive), although keeping in
agents and the search of new items can be done off-line. mind the restrictions mentioned above. The fact of having
Whenever a new item for CV is found, the CVSA asks implemented agents in different frameworks (the original
the MASplan MA for sending an e-mail to the agents were implemented in FIPA-OS tool) is not a
corresponding user, informing of the result of the search. problem as both frameworks are FIPA-compliant.
This way, the user can obtain a copy of the found data for Regarding to the use of ontologies by the agents, the
other purposes different to the generation of CVs. Apart original MASplan one has complemented with the
from this interaction, the CVSA send these data to the UA inclusion of new concepts related to CV activities. As
for its inclusion in the local user CV database. example of these new definitions, the following OWL
The implementation of these local databases does not code points that research activities related to the
affect the dynamic generation of the forms as the XSD's Philosophy area are disjoint with those related to
are accessed from a remote server. Biomedicine.
As stated in the introduction of the paper, a centralized
database could have been a good solution. Nevertheless, <owl Class
and in spite of the security offered by the database rdf:about="#Philosophy_Research Activity">
manager, a significant number of members of the

12
 <rdfs:subClassOf> changeuser.xml, portal.xsl and sunriseconfHTML.xsl.
<owl :Class rdf:about="#Research Activity"/> The data involved in the structure of the system have
</rdfs:subClassOf> been divided into four categories, declared in their
<owl :disjointWith> respective schemas.
<owl :Class rdf:about="#Biomedicine * Courses (lectures, seminaries): name of the course,
Research-Activity "/> recommended bibliography, etc.
</owl :disjointWith> * Personal data: name, address, spoken languages,
</owl :Class> etc.
* Items: list of the items to be included in a CV, such
This type of definitions makes easier the interaction as publications in journals, conferences, books,
among the CVSA's and the UA's. When both agents start chapters, research activities, etc.
a conversation, they compare the research areas of their * Groups: list of research groups, members, address,
corresponding users. In case of disjointness, there is no director, etc. - accessed only by the administrators
need of search in the XML database, avoiding its of the system.
computational cost, and maybe more important, avoiding It is noted that for avoiding repetition in this structure,
future unproductive conversations. A more refined each document includes a reference field, a kind of main
version, currently in progress, will consist in using the DF key in a database. This key is the identifier (Spanish DNI)
functionality as a yellow pages service of research of the user.
activities. As stated above, the web environment is generated in
Another open line in this project consists of providing a dynamic way. It is interesting that the information could
more intelligence to the system. The purpose would be change following the structure of the XML files, in other
that the CVSA's themselves were able to extract the rules words, following the changes produced in the XML
of interaction among users regarding analogue research schemas. This purpose has been reached through the
areas- for example, if two areas are compatible or not - design of a set of XSL transformations that allow
from the list of keywords in the titles of publications or generating Cocoon forms. As expected, a change in the
common authors in the items. For that purpose, several XML schema involves a change in the corresponding
well-known techniques can be used, e.g., Dempster-Shafer CForm.
method [10]. After that deduction, the CVSA would For this purpose, three transformation files were
interact with the RuA and OA for the inclusion of the designed: form-template, form-definition and form-
deduced axiom in the ontology. binding. These transformations are applied to the XSD
files in order to generate the forms presented to the user.
A possible problem could be that the application of three
4. Description of Cocoon Web portal transformations whenever a form is generated could be
excessively slow. Fortunately, Cocoon offers an effective
In this section, the Cocoon web portal will be cache system. When a form is loaded for the first time, the
described. It is noted, as stated above, that the interaction rest of forms can be showed in a more quick way. This
of the web portal with the MAS is done via the UA's. The methodology has been simultaneously used by other
Cocoon distribution offers a good number of useful developers, in particular, Arje Cahn and Max Pfingsthorn
developing examples. From one of these examples, in from Hippo.nl. However, that implementation is even
particular from the address samples/blocks/portal- much more limited than the presented in this work. As
fw/sunspotdemoportal, the authors have developed their example, the date type is treated as a mere string.
web environment. The mentioned files collect the XSD structure and
After an initial verification process, the user is extract recursively the data and turn it into a CForm
presented a frame configuration where several parts can structure. All of them have a similar structure, based on
be distinguished. The most important one is located on the recursive invocations with attribute inheritance and
left, including a set of controls. Clicking on the controls, a recollection in groups of information. Thus, the
user can insert, modify, look for and update their data, as differences are not in their structure but in their content,
generating several formats of CV. An example of due to their specific syntax using their respective
generated CV CForm is shown in Figure 2. It is important namespaces.
to remark the identifier is the parameter used to match the It is noted that for each transformation three possible
data with the corresponding user. This is necessary due to simple types have been considered:
it is clearly not desirable that a user could access data * Those defined as xs:simpleType.
related to other users. The use of this parameter in the
* Those defined as complex type, but they are
session is not trivial and requires modifying some XML
and XSL files, such as sunrise-newuser.xml, sunrise- xs:extension.

13
* Those types that have directly defined from the tags can be applied to a simple type. In the case of the
complex type, for example, a complex type with validation tags, they are restricted to xs:pattern,
attribute type= "xs:string". xs:minLength, xs:maxLength, xs:maxInclusive,
It has been observed that there are some fields that are xs :minInclusive.
rather difficult to manage. A significant case is that of the
"author" field in the file including the CV Items. This
field should include the number of the identify card of
each author involved in the item (for example, the list of
authors of a paper accepted for publication in a journal).
Nevertheless, these numbers are difficult to remember by
the users. It would be desirable an easier way for the
inclusion of the authors. This problem has been resolved
including a new attribute, called "src", to the tags
xs:element and xs:attribute, called "Combo-Professor" Apellidol:

when it is necessary. This name is related to the match _~~~~~~~

actions -in the sitemap file located in the directory forms-
that indicate that users should be inserted through a
combo widget instead of a textfield one. Iigure2T:Afdealo Eliffiii

As example of design of XSL transformation, the

authors will describe the transformation in order to get the Figure 2: A detail of generated CV form
form template file.

4.1 Design of the form template The first phase of the project was only focused on this
web environment. When a user desires to generate an
Mainly, the idea consists in reading the schema from updated CV, a XSL-FO transformation is applied in order
to get that CV which is accessed in the browser. The
the root element to its children elements, at the same time
as each child element inherits the attributes of its parent system has been used successfully for the generation of
element. This structure will be used whenever the system the documentation regarding the PhD studies program of
three departments of the University of La Laguna. That
needs to extract information from a XSD document.
program covers nearly 100 professors, thus the
recollection of information of the lectures and giving
4.2 Restrictions in the transformations homogeneity to the presentation of that information could
involve a great amount of time without that web
The designed transformations do not cover all the environment (as stated above, the mean time for the
possible XSD syntax. As it is said in Apache web site generation of a CV has been calculated as 3 hours).
Moreover, the users can get their CV with only a click
"The difficulty with schemas is that they can contain when it is required or generate their web page with their
an implicit structure by means of references to elements" research data.
However, due to the designed structure, it is possible
to make easily updates and code improvements. The main 5. Conclusions
restrictions in the XSD structure are summarized as
follows: In this paper, the authors have presented a multi-agent
There can be as many complex and simple types as system for the automatic generation of CVs for a
desired, but it is not allowed that the root element- or Universitary scenario. The agents related to this purpose
their children - has those types as children elements. have been integrated with a previously designed system,
Each immediately children element of the root called MASplan, for planning and scheduling in a
element must define an attribute "xs:ID", that identifies University Research Group Scenario. The included
univocally that element. The name of that ID element agents, called CVSA, are mobile- one for each user in the
must be the same for all the elements of the same system- and their task consists of the search of items
document. This element is used due to two reasons: related to their corresponding user. This search is carried
* Avoiding repeated id's. out interacting with other agents, called UA, in charge of
* Allowing the search of elements for the management of local CV item databases. The main
modifications and deletions. reason for this distribution, and thus for the use of a
Datatypes ,documentation, cardinality and validation multiagent system, lies in the human behaviour. Firstly, it

14
has been observed that sometimes the corresponding inteligentes: Aplicacion a la planificacion automaitica de
author of an article forgets to communicate the acceptance agendas y al control de procesos". Ph.D. Thesis,
of that paper to the rest of the authors and that a co-author University of La Laguna.
is usually so busy that he/she prefers to update his/her CV
database later, taking the risk of 'losing' the paper in [3] E.J. Gonzailez, A. Hamilton, L. Moreno, R.
his/her CV. Secondly, a significant number of members of Marichal, J. Toledo. (2004) "Ontologies in a Multi-Agent
the Universitary Scenario were reluctant to insert their System for Automated Scheduling". COMPUTING AND
data, in a centralized system. In this context, the features INFORMATICS, 23 (2). 157-178.
provided by the multiagent systems - distribution,
reliability, proactivity, autonomous and reactive [4] FIPA Abstract Architecture Specification. Technical
behaviour, etc.- seem to be especially useful. Report, SCOOOOIL. FOUNDATION FOR
The original MASplan system included an ontology INTELLIGENT PHYSICAL AGENTS. December, 2002.
for the automatic meeting negotiation in an intelligent way
amongst several researchers, among other several features. [5] S. Falasconi, G. Lanzola, M. Stefanelli. (1996) "Using
This ontology was implemented in OWL, a highly ontologies in Multi-Agent Systems". Tenth Knowledge
expressive markup language. In the context of this work, Acquisition for Knowledge-Based Systems Workshop.
some new concepts have been introduced, for example
axioms related to the disjointness of different research [6] Fonseca S.P., Griss M.L. and Letsinger R. (2002)
areas, facilitating the interaction among the agents "Agent Behavior Architectures. A MAS Framework
involved. Currently, the authors are working on the Comparison". Proceedings of the first international joint
addition of new interesting features in the system, as the conference on Autonomous agents and multiagent
ability of the agents of deducing axioms on-line. systems, Bologna, Italy.
Other interesting aspect of this work is the
implementation of a dynamic web environment for the [7] Stefano Mazzocchi, (2000) "Adding XML
management of the CV generation. The development of Capabilities with Cocoon", ApacheCON 2000 - Orlando,
this environment has been carried out using Apache 2000
Cocoon, a mature and popular web development
framework, developed in Java and based on the Servlet [8] Introducing Apache Cocoon", Available:
API model. The web environment is generated in a http://cocoon.apache.org/2.1/introduction.html
dynamic way. The information changes following the
structure of the XML files, in other words, following the [9] Steven Noels. (2003) "Standards Applied: Using
changes produced in the XML schemas. This purpose has Apache Cocoon and Forrest", XML Europe 2003,
been reached through the design of a set of XSL London, England.
transformations that allow to generate Cocoon forms. As
expected, a change in the XML schema involves a change [10] Bauer M. (1995), "A Dempster-Shafer Approach to
in the corresponding CForm. Modelling Agent Preferences for Plan Recognition", User
The system has been used successfully for the Model. User-Adapt. Interact. 5(4): 317-348.
generation of the documentation regarding the PhD
studies program of three departments of the University of
La Laguna, Spain. That program covers nearly 100
professors, so the use of this web environment has saved a
lot of time in the data treatment.

10. References
[1] E.J. Gonzailez, A. Hamilton, L. Moreno, R.
Marichal, S. Torres. (2003) "Masplan: A Multi-Agent
System for Automated Planning and Scheduling in a
University Research Group Scenario". 1st
Multidisciplinary International Conference on Scheduling:
Theory and Applications. Nottingham.

[2] E.J. Gonzalez, (2004) "Diseno e implementacion de

una arquitectura multiproposito basada en agentes

15