Dynamic Access Control Framework for Enterprise Content

Management Systems

Nadia Hocine and Ismail Bokhari

University of Abdelhamid Ibn Badis of Mostaganem, Mostaganem, Algeria

Keywords:

Access Control, Enterprise Content Management System.

Abstract:

With the large adoption of telework business model, employees can work anytime, anywhere and sometimes

with their own personal devices due to the limited ﬁnancial capabilities of their companies. Many issues of

the security and access control of remote exchanges of Enterprise Content Management systems (ECM) have

to be considered. In particular, the access control should be adapted to employees’ context, their multiple

device capabilities as well as their proﬁles and situations to increase the usability of the system. However,

most access control models do not take into account the users’ proﬁles and the variability of their devices in

open network. They also focus on the continuous intervention of administrators to manage the system and

add new devices and set parameters. With the diversity of users’ devices and context conditions in telework,

access control needs to be dynamically managed to reduce human intervention. In this paper, we suggest an

agent-based access control framework that focuses on M5StickC external device used as an access badge. The

framework is based on a multi-level rule engine to dynamically generate policies according to users’ context,

proﬁle and device. It is implemented and proposed as an open-source solution for small companies to manage

their own ECM access control.

1 INTRODUCTION

Telework becomes a popular alternative to work that

consists in relocating work to reduce costs while in-

creasing the well-being and the safety of employees.

It may also be an alternative solution for employees

with different proﬁles and high mobility. Managing

remotely enterprise resources and their access can be

challenging when employees work with their personal

devices, under mobility condition (such as home, air-

port and train) and with particular disabilities. There-

fore, the access control management should take into

account not only the context and user various devices

but also their proﬁle in terms of work and health con-

ditions in order to enhance the usability of the access

control applications.

Diffrent dynamic access control strategies have

been proposed in the literature to deal with dynamic

nature of user environments and contexts (Freuden-

thal et al., 2002) (Calo et al., 2018) (Oluwatimi et al.,

2018). However, most access control models do not

take into account the variability of users devices in

open network as well as usability issues. Moreover,

most strategies require human administrator interven-

tion (Calo et al., 2018) which can be a challenging

task, especially in the case of variability of devices

and users’ proﬁle.

In this paper, we suggest a dynamic access con-

trol framework that focuses on an agent-based system.

Providing users with an intelligent system can im-

prove not only the management of contextual data for

access control in a lightweight decentralized infras-

tructure (Uddin, 2019), but also allows more person-

alized access control that deals with usability issues.

Indeed, with their capability to think and make deci-

sion, software agents can control the dynamic context

information and can help in making decision on re-

sources access authorization. Agents use a rule-based

system to set resources permissions. In fact, an ac-

cess control system is based on the expert decisions

on knowledge constituted from various data on users

and their context information that can be represented

using rules (B

adic

a et al., 2011). However, a rule-

based system may require high level computing per-

formance with a large number of rules, known as the

redundancy issue (B

adic

a et al., 2011). To deal with

this issue, our proposed framework considers a multi-

level rule engine that selects only a set of rules to gen-

erate dynamically an access control policy on the ba-

sis of resources’ sensibility.

The rest of this paper is organized as follows: Sec-

tion 2 discusses related works on dynamic access con-

Hocine, N. and Bokhari, I.

Dynamic Access Control Framework for Enterprise Content Management Systems.

DOI: 10.5220/0010514806470652

In Proceedings of the 18th International Conference on Security and Cryptography (SECRYPT 2021), pages 647-652

ISBN: 978-989-758-524-1

647

trol techniques. In section 3, we introduce our pro-

posed agent-based control framework. We describe

in this section the architecture of the proposed frame-

work as well as the policy speciﬁcation and genera-

tion. Section 4 presents the implementation details

and the test of the proposed framework. Finally, Sec-

tion 5 concludes the paper and outlines our future

works.

2 RELATED WORKS

Access control is the veriﬁcation of whether user ac-

tions on resources are authorized according to a secu-

rity policy (Sandhu and Samarati, 1994). It allows

controlling the users’ access to sensitive resources

(Nyakomitta and Abeka, 2020). Different access con-

trol approaches have been proposed in the literature

according to speciﬁc systems.

Role-based access control approach has been ini-

tially adopted in various institutions and companies

due to its simplicity in large-scale authorization man-

agement (Ferraiolo et al., 2016). This approach uses a

model that considers three concepts: users, roles and

permissions. Roles are users’ job functions within

the organizations and permissions are the approvals

to perform certain operations on resources. The ac-

cess permissions are assigned to roles in order to man-

age easily permissions (Liu et al., 2017). However,

the role based policy model does not consider context

constraints such as user location and devices.

Dynamic access control approaches can be used

to deal with the ﬂexible control of resources ac-

cess permissions according to available context in-

formation (Corradi et al., 2004). For instance, dif-

ferent works extended the role-based approach by

considering temporal and spatial information (Joshi

et al., 2005) (Damiani et al., 2007) as well as the

resource and environment dimensions (Hosseinzadeh

et al., 2016) (Trnka and Cerny, 2016). They fo-

cused on the dynamic user-role and role-permission

assignments through contextual constraints, called

dynamic attributes (Zheng et al., 2011). For exam-

ple, Kulkarni and Tripathi proposed a programming

framework that extended the role-based access con-

trol model by considering user and resource condi-

tions in role-permission assignments (Kulkarni and

Tripathi, 2008). In Trnka and Cerny’ work, user roles

and context information such as the IP addresses and

times of the day were used to determine resources au-

thorization (Trnka and Cerny, 2016). In addition to

temporal and spatial constraints, access control can

also consider different attributes and real-time user’s

situations. For instance, Kayes et al. proposed a

situation-aware access control framework based on an

ontology (Kayes et al., 2019). A policy model was

deﬁned using conditional expression of users’ situa-

tions. In the previous reviewed works, the user-role

assignment policies related to dynamic context con-

straints require the continuous administrator interven-

tion. They also depend on context data regardless

users’ proﬁles and their devices capabilities and vari-

ability.

With the technological advancement of devices

and their embedded sensors, various approaches were

based on users’ devices to control their own access

to resources (Squicciarini et al., 2009) (Calo et al.,

2018). For instance, Verma et al. suggested a frame-

work in which the manager uses an interaction graph

to deﬁne the roles that devices may play in the sys-

tem (Verma et al., 2017). Other works considered

the networked environment model in order to give

the possibility of each device to obtain security con-

straints of other devices of its environment (Squiccia-

rini et al., 2009). However, this approach requires the

deployment of an environment model speciﬁcation

that should be supervised by administrators which can

be, in turn, a challenging task.

The previous reviewed access control models are

often based on centralized management of permis-

sions by an authorization entity which can reduce its

performance with the increase of number of users and

requests. Existing dynamic access control models can

difﬁcult to execute in distributed systems that have a

large number of users and/or ressources. The central-

ized management of access control in distributed sys-

tems may cause access delay or unauthorized access

because of the server overload. Moreover, due to its

centralized nature, it can become easily the target of

network malicious attacks.

To deal with the previous issues, a distributed

management of access control can be used. It consists

in allowing each node of the network to control the

dynamic mapping between users and permissions to

improve the access control management performance.

It can be also used to personalize the access control

to users with various devices, proﬁles and situations

without requiring continuous administrator interven-

tion. In this paper, we propose an agent-based ac-

cess control framework for Entreprise Content Man-

agement systems (ECM). Agent-based systems can

play an important role in building distributed intelli-

gent access control management frameworks that take

into account various user device, context and proﬁles.

Unlike dynamic access control models that not con-

sider open network due to vulnerability issues and

excessive intervention of administrators, we suggest

a distributed access control management that focuses

SECRYPT 2021 - 18th International Conference on Security and Cryptography

648

on an agent-based architecture. We also provide users

with an external ”access badge” to initiate and obtain

their access permissions. The objective is to deal with

the variability of user devices and their limited capa-

bilities in terms of energy and sensors disposition. In

order to limit vulnerabilities in the case of mobility,

the access request is limited to Bluetooth connexion

with small amplitude parameters and within limited

time. Next, we describe the proposed framework as

well as the dynamic policy generation for ECM.

3 AN AGENT-BASED ACCESS

CONTROL FRAMEWORK FOR

ECM SYSTEMS

Figure 1: General framework.

In this paper, we suggest an access control framework

for enterprise content management systems (ECMs).

ECM is a software program that consists of processes,

procedures, and technologies used in conjunction to

manage unstructured content such as documents and

ﬁles of enterprise (Katuu, 2018). It helps employees

upload resources that are identiﬁed by their meta-data

attributes. The latter are used in access control and

can be classiﬁed according to: resource owners, re-

source type and category as well as resource sensitiv-

ity.

With the rapid evolution of telework adoption in

our society nowadays, employees may use various

personal devices to communicate and get services of

the ECM anywhere and anytime. Different contex-

tual information, such as user location, request time

and mobility can be used to enforce access control

and system security. Contextual information are con-

trolled through independent services and programs

that act as intelligent agents and help the develop-

ment of more intelligent access control systems (Ud-

din, 2019) (Calo et al., 2018).

The proposed agent-based access control frame-

work is dedicated to small companies that need the

personalization of access control by taking into ac-

count users’ context and conditions while limiting ad-

ministrator intervention. As shown in Figure 1, the

user is provided with a M5STICK device used to ob-

tain access authorization from user device (such as

laptop and phone). M5STICK device includes various

features and sensors such as Bluetooth 4.0, Wiﬁ, Ac-

celerometer, etc. Using his or her M5STICK badge,

an employee can acquire permissions with a Wiﬁ or

Bluetooth connection to his or her telework device.

We used Bluetooth connection in this work. Indeed,

a Bluetooth-based infrastructure makes the system

more resilient to malicious attacks especially in user

mobility conditions (Wong and Hunter, 2017).

We used Arduino development environment in or-

der to manage Bluetooth connection with devices and

to obtain contextual data. After receiving the access

request, the authorization server will be responsible

on generating the policy using a rule-based system.

Then, once the permission is accorded, the user can

obtain access to resources using the ECM application

server. We used in this paper Alfresco bitnami VM

(Pal, 2016) as an example of ECM system. Next, we

describe the architecture of the framework as well as

the policy speciﬁcation and generation.

3.1 Access Control Framework

Architecture

The framework focuses on an agent-based architec-

ture (see Figure 2). Agents are software entities that

can send and receive data from sensors or from other

agents in a particular environment. According to

Russell ”An agent is anything that can be viewed as

perceiving its environment through sensors and act-

ing upon that environment through effectors” (Rus-

sell and Norvig, 2002). In the proposed framework,

agents are used to represent software entities that con-

trol the user contextual data as well as to ensure the

internal functioning of the access control through pol-

icy generation. It consists of: the user agent, autho-

rization server that includes an agent policy generator

as well as an authoring tool.

• User Agent. It is responsible on communicating

with the authorization server to acquire the ac-

cess policy. It also provides the latter with context

information from sensors. In fact, to obtain per-

missions on his or her laptop or smartphone, the

user has to obtain access authorization through a

M5STICK badge. Once the access is permitted,

the user can communicate with the ECM applica-

tion server to get authorized resources using his

own User Agent.

Dynamic Access Control Framework for Enterprise Content Management Systems

649

Figure 2: Framework architecture.

• Authorization Server. It aims to control dynam-

ically the user permissions on resources by gen-

erating appropriate policy and providing it to the

user. It consists of the following elements: rule

based system to represent the expert decisions on

knowledge constituted from various data on users

and their context information, the authoring tool

used by the administrator to introduce general per-

mission constraints and the agent policy genera-

tor that receives users’ access request though their

User Agents and generates policies.

• The Application Server. It allows performing ac-

tions on resources of the enterprise content man-

agement system. The access permissions ac-

corded to the user follow the policy generated by

the authorization server.

• User Proﬁle. Each user of the system is charac-

terized by a proﬁle which represents a set of at-

tributes related to the user such as: age, mobility,

disabilities, kind of contract, etc. It is used to per-

sonalize access control constraints to user speciﬁc

conditions.

3.2 Policy Generation

Policy speciﬁcation may depend on different lan-

guages. For example, the role-based access control

policy may be expressed using access control lists

(ACL) to resources stored in a central or distributed

memory (Verma et al., 2017). It is a simple way to

map resources to permissions while considering re-

sources hierarchy. However, one of the limitations of

ACLs is the low expressiveness of regular expressions

for dynamic contextual data and meta-data based re-

sources identiﬁcation.

XML (extensible markup language) based policy

speciﬁcation has been largely used due to its extensi-

ble nature and simplicity to specify policy expression

with different data types. Extensible access control

markup language or XACML was proposed initially

for role-based approach and has then been extended

and used in various works (Ferraiolo et al., 2016).

Using XACML to encode the policy consists in the

deﬁnition of expressions that return values that reﬂect

the kind of authorization or deny of access.

In this paper, we focus on XML-based policy lan-

guage inspired by the XACML speciﬁcations. Un-

like existing policy generation techniques, we focus

on a rule-based generation that depends on each de-

vice and user attributes. It also can be initiated by the

administrator by introducing only general rules using

an authoring tool that does not require speciﬁc techni-

cal skills. The policy XML ﬁle generation is based on

expert knowledge through a rule based system to infer

the decision on resources permission. Using Markup

Scheme such as XML also allows the communication

between agents and data transfer via a network.

Figure 3: Policy speciﬁcation.

The Agent Policy Generator uses the rule-based sys-

tem to generate a policy by mapping user-agents to

permissions. However, a rule-based system may re-

quire high level computing performance due to the

large number of rules, known as the redundancy is-

sue. We propose therefore a personalized access con-

trol that reduces the number of rules used to generate

dynamically a policy. Indeed, the Agent Policy Gen-

erator selects only a small set of eligible rules accord-

ing to a security level that depends on resources sen-

sibility. For instance, according to the resources low

level of sensibility, the agent can exclude for instance

rules that consider the context information, such as

working time and location.

According to resources sensibility, three rules lev-

els were considered in our framework. The ﬁrst level,

consists of role-based rules that represent the con-

straints that use agent-role mapping without consider-

ing context and proﬁle information. The second level

is based on user proﬁle data such as availability and

health conditions. The third level considers the con-

textual information through the dynamic environment

characteristics mapping with resources classiﬁcation.

SECRYPT 2021 - 18th International Conference on Security and Cryptography

650

The Policy Generator Agent generates an XML ﬁle

with the appropriate policy that uses the general spec-

iﬁcation shown in Figure 3.

4 IMPLEMENTATION

The project repository can be obtained from Github,

with this link: https://github.com/AuthorisationTool

We have performed an initial test of the framework

using M5StickC device and two laptops. The devices

used were Intel I7-7500U, 8 GB DDR4 RAM, SSD

240 GB. M5StickC device is an internet of things

development card based on ESP32 micro-controller

used in our framework as an access badge. It con-

sists of various modules especially: Bluetooth 4.0,

Wiﬁ, Gyro-meter, Accelerometer, Microphone, USB,

IR Transmitter, LCD 0,96” screen and 3 Buttons.

Each badge characteristics are stored in a Post-

greSQL database of the authorization server. This in-

cludes the identiﬁer of the badge and the identiﬁer of

the process HID (Holder ID) used to identify the user.

Using a Bluetooth 4.0 connection, the user can obtain

access from his or her telework device. The server of

the badge was based on GATT protocol and Google

Chrome as navigator. Figure 4 shows the badge used

in our ﬁrst test.

Figure 4: M5StickC device.

Figure 5: Example of generated policy.

To generate policies, rules were implemented

through a simple rule engine that we developed. An

example of generated policy is shown in Figure 5. The

proﬁle constraint that we have considered in our ﬁrst

test was: ”Health” which is a list of abstract health

conditions of the user such as ”Healthy” and ”Dis-

ability”. The latter determines personalized time and

spatial conditions to be considered for speciﬁc user’s

health conditions. As for the context data, three con-

straints have been included in the test : (i) ”User

Agent location” which represents the geographical

site position that includes values on latitude, longi-

tude and diameter of user location (ii) ”Gyro” which

considers user movement data that includes the speed

recorded when the user asks access permission (iii)

”Acceleration” of the agent movement.

We developed an authoring tool that help admin-

istrators easily set and update the general rules of

access control without requiring high development

skills. Using the authoring tool Web interface, the

administrator set ﬁrst the possible ressources types,

category, sensitivity levels and owners for role cre-

ation. The ressources conﬁguration follows the possi-

ble considerations of Alfresco ECM. Once the role

is created, the administrator can add levels that in-

clude context and/or proﬁle constraints. The adminis-

trator can also add his or her personalized constraints

or uses the default possible constraints.

The results of the ﬁrst test with two experts

showed the simplicity of the interface and efﬁciency

of the access badge to get rapid and easy access to the

Alfresco server application.

5 CONCLUSION

In this paper, we presented an agent-based access

control framework architecture for enterprise con-

tent management systems. An open source tool is

proposed following this architecture to help small

companies in controlling access to ECM ressources

while limiting administrator’s intervention. The tool

is based on a rule engine that generates dynamically

an access control policy on the basis of users’ condi-

tions and context information obtained with the help

of an M5StickC access badge. Furthermore, we pro-

posed an authoring tool that allows administrators to

introduce general constraints without having strong

knowledge about the rule and policy speciﬁcation.

The results of the initial test of the tool with two ex-

perts showed the simplicity of the interface and efﬁ-

ciency of the access badge to get rapid and easy ac-

cess to the Alfresco server application. In our future

works, we plan to test the performance of this tool

Dynamic Access Control Framework for Enterprise Content Management Systems

651

with a large number of employees and administra-

tors through different case studies. The objective is

to evaluate the efﬁciency of the tool to improve both

authoring tool and policy generation. Finally, we plan

to test the usability of the system with employees with

particular health conditions (such as post-stroke pa-

tients) by considering various proﬁle-based rules that

include personalized spatial and temporal constraints.

REFERENCES

adic

a, C., Braubach, L., and Paschke, A. (2011). Rule-

based distributed and agent systems. In International

Workshop on Rules and Rule Markup Languages for

the Semantic Web, pages 3–28. Springer.

Calo, S., Verma, D., Chakraborty, S., Bertino, E., Lupu, E.,

and Cirincione, G. (2018). Self-generation of access

control policies. In Proceedings of the 23nd ACM on

Symposium on Access Control Models and Technolo-

gies, pages 39–47.

Corradi, A., Montanari, R., and Tibaldi, D. (2004). Context-

based access control for ubiquitous service provision-

ing. In Proceedings of the 28th International Com-

puter Software and Applications Conference, pages

444–451. IEEE.

Damiani, M. L., Bertino, E., Catania, B., and Perlasca, P.

(2007). Geo-rbac: a spatially aware rbac. ACM Trans-

actions on Information and System Security, 10(1).

Ferraiolo, D., Chandramouli, R., Kuhn, R., and Hu, V.

(2016). Extensible access control markup language

(xacml) and next generation access control (ngac).

In International Workshop on Attribute Based Access

Control, pages 13–24.

Freudenthal, E., Pesin, T., Port, L., Keenan, E., and Karam-

cheti, V. (2002). drbac: distributed role-based ac-

cess control for dynamic coalition environments. In

Proceedings 22nd International Conference on Dis-

tributed Computing Systems, pages 411–420. IEEE.

Hosseinzadeh, S., Virtanen, S., D

ıaz-Rodr

ıguez, N., and

Lilius, J. (2016). A semantic security framework and

context-aware role-based access control ontology for

smart spaces. In Workshop on Semantic Big Data,

pages 1–6.

Joshi, J. B., Bertino, E., Latif, U., and Ghafoor, A. (2005). A

generalized temporal role-based access control model.

IEEE transactions on knowledge and data engineer-

ing, 17(1):4–23.

Katuu, S. (2018). A comparative assessment of enter-

prise content management maturity models. In E-

manufacturing and e-service strategies in contempo-

rary organizations, pages 93–118. IGI Global.

Kayes, A., Han, J., Rahayu, W., Dillon, T., Islam, M. S.,

and Colman, A. (2019). A policy model and frame-

work for context-aware access control to information

resources. The Computer Journal, 62(5):670–705.

Kulkarni, D. and Tripathi, A. (2008). Context-aware role-

based access control in pervasive computing systems.

In Access control models and technologies, pages

113–122.

Liu, Q., Zhang, H., Wan, J., and Chen, X. (2017). An access

control model for resource sharing based on the role-

based access control intended for multi-domain man-

ufacturing internet of things. IEEE Access, 5:7001–

7011.

Nyakomitta, P. S. and Abeka, S. O. (2020). Security inves-

tigation on remote access methods of virtual private

network. Global Journal of Computer Science and

Technology.

Oluwatimi, O., Damiani, M. L., and Bertino, E. (2018). A

context-aware system to secure enterprise content: In-

corporating reliability speciﬁers. Computers & Secu-

rity, 77:162–178.

Pal, V. (2016). Alfresco for Administrators. Packt Publish-

ing Ltd.

Russell, S. and Norvig, P. (2002). Artiﬁcial intelligence: a

modern approach.

Sandhu, R. S. and Samarati, P. (1994). Access control: prin-

ciple and practice. IEEE communications magazine,

32(9):40–48.

Squicciarini, A. C., Shehab, M., and Paci, F. (2009). Col-

lective privacy management in social networks. In

Proceedings of the 18th international conference on

World wide web, pages 521–530.

Trnka, M. and Cerny, T. (2016). On security level usage in

context-aware role-based access control. In Proceed-

ings of the 31st ACM Symposium on Applied Comput-

ing, pages 1192–1195.

Uddin, I. (2019). A rule-based framework for developing

context-aware systems for smart spaces. PhD thesis,

University of Nottingham.

Verma, D., Calo, S., Chakraborty, S., Bertino, E., Williams,

C., Tucker, J., and Rivera, B. (2017). Generative pol-

icy model for autonomic management. In 2017 IEEE

SmartWorld, Ubiquitous Intelligence & Computing,

pages 1–6. IEEE.

Wong, K. and Hunter, A. (2017). Bluetooth for decoy sys-

tems: A practical study. In 2017 IEEE Conference on

Communications and Network Security (CNS), pages

86–387. IEEE.

Zheng, J., Zhang, Q., Zheng, S., and Tan, Y. (2011). Dy-

namic role-based access control model. Journal of

software, 6(6).

SECRYPT 2021 - 18th International Conference on Security and Cryptography

652