WebC-Docs

A CMS-based Document Management System

ao de Sousa Saraiva and Alberto Rodrigues da Silva

INESC-ID & SIQuant, Rua Alves Redol, 9, 1000-029 Lisboa, Portugal

Keywords:

Content Management System, Document Management System, WebComfort.

Abstract:

Content Management Systems (CMS) are typically regarded as critical software platforms for the success

of organizational web sites and intranets. Nevertheless, a simple CMS alone does not provide enough sup-

port for a typical organization’s requirements, such as document management and storage. On the other hand,

Enterprise Content Management (ECM) systems are typically regular web-applications that provide such func-

tionality but without the added advantage of being based on a component-based CMS platform.

This paper presents the architecture and major technical details of WebC-Docs, a highly-customizable toolkit

(for the WebComfort CMS platform) that provides document management functionality. Due to this, the

WebC-Docs toolkit can be conﬁgured and used in various kinds of scenarios.

1 INTRODUCTION

The Internet’s recent expansion has led to the appear-

ance of many web-oriented CMS (Content Manage-

ment Systems) (Boiko, 2001) and ECM (Enterprise

Content Management) (Kampffmeyer, 2006; Rock-

ley, 2002) platforms with the objective of facilitating

the management and publication of digital contents.

CMS systems can be used as support platforms for

web-applications to be used in the dynamic manage-

ment of web sites and their contents (Carmo, 2006;

Saraiva and Silva, 2008). These systems typically

present aspects such as extensibility and modularity,

independence between content and presentation, sup-

port for several types of contents, support for access

management and user control, dynamic management

of layout and visual appearance, or support for work-

ﬂow deﬁnition and execution. On the other hand,

ECM systems are typically regular web-applications

oriented towards using Internet-based technologies

and workﬂows to capture, manage, store, preserve,

and deliver content and documents in the context of

organizational processes (AIIM, 2009). These two

content-management areas are not disjoint, and it is

not unusual to ﬁnd a CMS system acting as a repos-

itory for an organization’s documents and contents,

albeit at a very “primitive” level (e.g., no checking

for duplicate information, no grouping of documents

according to a certain logical structure, and no possi-

bility of providing metadata for each document).

Into this context comes the WebC-Docs system,

a document management toolkit for the WebComfort

CMS platform (WebComfortOrg, 2009) providing a

large set of conﬁguration points that allow the system

to be used in a wide variety of application scenarios.

In this paper we present the architecture of WebC-

Docs and its major technical contributions. This pa-

per is structured as follows. Section 1 introduces

CMS and ECM systems, and the classical problem of

document management in an organizational context.

Section 2 presents WebC-Docs’ architectural aspects.

Section 3 provides a brief discussion of some addi-

tional issues regarding this system. Section 4 presents

related work that we consider relevant for this project.

Finally, Section 5 presents the conclusions so far, as

well as future work.

2 WebC-Docs

WebC-Docs is a document management toolkit

for the WebComfort CMS platform (Saraiva and

Silva, 2008; WebComfortOrg, 2009), featuring a

component-based architecture that makes it easily ex-

tensible, highly conﬁgurable, and adequate for several

kinds of application scenarios.

This section presents some relevant components

and aspects of WebC-Docs’ architecture, namely: (1)

its main concepts; (2) its integration with the Web-

Saraiva J. and da Silva A. (2009).

WebC-Docs - A CMS-based Document Management System.

In Proceedings of the International Conference on Knowledge Management and Information Sharing, pages 21-28

DOI: 10.5220/0002287500210028

 SciTePress

Comfort CMS platform; (3) its Explorer-like web in-

terface; (4) the Dynamic Attributes mechanism; (5)

the Indexing and Searching functionality; (6) the Per-

missions mechanism; and (7) the facilities for using

additional repositories to store and locate documents.

2.1 Main Concepts

The WebC-Docs system consists primarily of the fol-

lowing concepts, also presented in Figure 1: (1)

Document; (2) Folder; (3) Document Shortcut;

and (4) Document and Folder Permissions.

Figure 1: The main concepts of the WebC-Docs system.

A Document is the main entity of WebC-Docs; in a

typical usage scenario, it represents a “real-world”

document (the document’s digitized ﬁle may or may

not be included in the Document; WebC-Docs allows

the existence of Documents as simple “placeholders”

for the real documents that they represent). On the

other hand, a Folder consists of a Document con-

tainer, but it can also contain Document Shortcuts,

which are mere “pointers” to regular Documents.

Thus, although a Document must be located in exactly

one Folder, it is possible for it to be accessed from

other Folders, by using Document Shortcuts.

Finally, Document and Folder Permissions are

associated with Documents and Folders, respec-

tively, and specify what actions each WebComfort

role can perform over them (the Permissions mech-

anism is explained further down this section).

2.2 Integration with WebComfort

One of the important innovations in WebC-Docs is

its integration with a CMS platform, in this case We-

bComfort (Saraiva and Silva, 2008). This integra-

tion consists mainly of the following points: (1) the

user’s main interaction points with WebC-Docs are

implemented as WebComfort modules (e.g., Docu-

ment Management, Statistics) that take advantage of

the facilities provided by WebComfort; (2) the Per-

missions mechanism (described below) uses Web-

Comfort roles and users, instead of (re)deﬁning those

concepts again; and (3) WebC-Docs is distributed as

a WebComfort toolkit, and so it can installed on any

WebComfort installation in a simple and automatic

manner. Figure 2 shows an overview of WebC-Docs’

integration with WebComfort and other toolkits (be-

cause of space limitations, this integration will not be

presented in detail in this paper).

Figure 2: An overview of the integration with WebComfort.

Besides all this out-of-the-box functionality pro-

vided by the CMS, an added advantage of such an

integration is that the system can be easily adapted to

requirements such as those derived from an organiza-

tion’s structure or size (e.g., an organization’s docu-

ment management portal can consist of a select num-

ber of tabs – searching, management, conﬁguration –,

or it can provide a tab for each of its users, each tab

containing the document management modules that

are adequate for the user’s responsibilities).

It is important to note that, although WebC-Docs’

current implementation is based on the WebComfort

CMS platform, it does not use concepts exclusive to

WebComfort (e.g., tabs, modules, and roles are con-

cepts that can be found in many CMS platforms).

Thus, WebC-Docs could easily be migrated to other

CMS platforms, as long as they provide adequate sup-

port for CMS-based web-applications.

2.3 Explorer-like Web Interface

One of WebC-Docs’ main goals is to be intuitive to

the “average” user. To achieve this goal, we designed

the main Document Management module to be sim-

ilar to the typical ﬁle explorer interface (such as Mi-

KMIS 2009 - International Conference on Knowledge Management and Information Sharing

crosoft’s Windows Explorer), as shown in the screen-

shot in Figure 3: the left side of the module displays

the system’s folder structure (always considering the

current user’s permissions), while the right side shows

the documents within the selected folder (if any).

Figure 3: The main interface: folders and documents.

The user can perform various actions over each

folder and document (such as “Delete”, “Move to”,

“Create shortcut to”, and “Export to ZIP ﬁle”), which

are displayed over the folder and document listings,

respectively. Documents can also be downloaded im-

mediately (by clicking on the ﬁle type’s icon), or the

document’s details page – illustrated in Figure 4 – can

be shown (by clicking on the document’s name).

Figure 4: The Document Details page.

WebC-Docs also provides the Recycle Bin (not

to be confused with Microsoft Windows’ own Recy-

cle Bin), to which deleted documents will be moved.

It will be up to a user with the “Documental Man-

ager” role (which can be conﬁgured to be any one of

WebComfort’s roles) to regularly check the Recycle

Bin and purge it, or restore documents that have been

deleted by mistake. The Recycle Bin can be disabled,

although this is generally not recommended, as delet-

ing a document would become an irreversible action.

2.4 Dynamic Attributes

One of WebC-Docs’ most powerful features is the

possibility of specifying metadata, in a customizable

manner that can be adjusted to the organization’s doc-

ument information requirements. This mechanism,

designated Dynamic Attributes, is based on the no-

tions of Attribute Set, Attribute Definition,

and Attribute Type, illustrated in Figure 5.

Figure 5: The “Dynamic Attributes” mechanism.

An Attribute Set consists of a group of

Attribute Types, and possibly even other

Attribute Sets, allowing the speciﬁcation of

metadata using a tree-like structure. Attribute

Type is responsible for deﬁning the various kinds

of attributes that be used (e.g., integers, dates,

enumerations). This is done by pointing the attribute

type to a class that implements an interface with

which WebC-Docs will communicate: this class will

provide the controls for viewing and editing attribute

values, as well as conversion to/from the type’s string

representation. The associations between Attribute

Sets and Attribute Types are called Attribute

Definitions, and they conﬁgure the Attribute

Types in the context of the Attribute Set (e.g.,

name, default value, whether it is read-only).

A user can apply any Attribute Set to any Doc-

ument or Folder, any number of times (e.g., to pro-

vide various author contacts for a certain Document).

Attributes are used to store user-provided values.

2.5 Indexing and Searching

For a document management system to be of any use,

it must allow its users to ﬁnd documents given only

some information about them (typically a string that

is a part of the wanted document). WebC-Docs pro-

vides a module with search functionality, using the

Lucene.Net indexing engine (LuceneNet, 2009), over

WebC-Docs - A CMS-based Document Management System

all document information (i.e., regular document at-

tributes, dynamic attributes, and document ﬁle con-

tents). There are two types of search: (1) “regular

search”, which operates like common search engines

(if any part of a document matches the user’s search

terms, the document is part of the search results); and

(2) “advanced search”, in which the user can specify

a variety of options (e.g., ﬁlter by dates, ﬁnd all doc-

uments with a certain string in their ﬁle contents) in

order to reﬁne the search and reduce the number of

false positives that will be returned by it.

Figure 6 presents the main components of WebC-

Docs’ ﬁle indexing. Whenever a Document is cre-

ated/altered, it is supplied as input to the Indexer

Front Controller, which quickly analyzes the

Document (namely its MIME type and ﬁle exten-

sion) and forwards it to any adequate Base Content

Indexers. Those indexers will then parse the Doc-

ument’s ﬁle contents and invoke WebC-Docs’ inter-

nal API (which, in turn, provides a wrapper around

Lucene.Net). Document metadata (regular document

attributes and dynamic attributes) is always indexed,

so the Document can still be found in the system.

Figure 6: Indexing support for many ﬁle formats.

WebC-Docs also allows searching in the Docu-

ment’s own repository (if the repository allows it,

of course), by means of the Additional repositories

mechanism, described further down this section.

We consider it relevant to note that, although

this indexing and searching functionality uses

Lucene.Net, the user’s search terms are not directly

provided as input to Lucene.Net. Instead, WebC-

Docs deﬁnes its own textual search language (simi-

lar to Lucene.Net’s and Google’s own search engine),

and uses a search term compiler built using CSTools

(CSTools, 2009). This compiler is used to both vali-

date the user search terms (e.g., too many parenthesis)

and generate an AST (Abstract Syntax Tree) that will

be provided as input to WebC-Docs’ internal search-

ing API (which will, in turn, transform that AST into

a Lucene.Net query string, and use that string to query

the various indexes that WebC-Docs uses).

This intermediate search step provides two main

advantages. First, it allows us to overcome some lim-

itations of Lucene.Net, such as: (1) some efﬁciency

issues when performing certain kinds of search (e.g.,

search without knowledge of the term’s preﬁx charac-

ters), or (2) its inability to search for all documents ex-

cept those that match a certain criteria (e.g., a search

string like “NOT document” in WebC-Docs would re-

turn all Documents that do not contain the word “doc-

ument” in them, but this search string would be con-

sidered invalid in Lucene.NET). Second it provides

the added advantage of additional pre-processing over

the user search terms (e.g., to remove potentially dan-

gerous words). Another possible advantage would be

that, if we later decided to use a different indexing en-

gine, the search language used by WebC-Docs’s users

would not need to be changed.

2.6 Permissions

Permissions speciﬁcation, which is perhaps the

most important requirement in this kind of appli-

cation, is addressed by WebC-Docs through the

concepts of Document Permissions and Folder

Permissions. This permissions mechanism follows

the typical ACL (Access Control List) philosophy,

in which each of a user’s operations (even viewing)

performed over a document/folder are ﬁrst checked

for the user’s permissions, to ensure that the user

can perform that operation over the speciﬁed docu-

ment/folder. To simplify the system’s usability, per-

missions are inherited (from parent folder to child

folder/document) by default.

The system can be conﬁgured to interpret permis-

sions in either the optimistic or strict perspectives. In

the optimistic perspective, access is granted to a user

over a certain folder/document if any of the user’s

roles has an explicit “allow” permission to it. On the

other hand, the strict perspective follows a more tra-

ditional approach, by blocking access if any of the

user’s roles has permissions explicitly stating that ac-

cess to the speciﬁed folder/document is blocked.

Figure 7 presents a folder’s permissions editing

screen: for each of the roles deﬁned in the WebCom-

fort installation, a large variety of permission options

can be conﬁgured (permissions for anonymous users

are also supported, if the organization wishes to make

some information publicly available).

Also, considering that these permissions are likely

to be accessed on a very frequent basis, WebC-

Docs uses a “permissions proxy” mechanism. This

KMIS 2009 - International Conference on Knowledge Management and Information Sharing

Figure 7: The permission options for a folder.

proxy stores information about accessed permissions

in the local server’s memory, which accelerates future

lookups of those permissions (e.g., for a user’s access

to a certain popular document). We say that this is a

proxy (and not just a simple cache) because all per-

mission modiﬁcations go through this proxy, which

removes the need to invalidate/remove entries (we just

need to update the proxy’s information).

2.7 Additional Repositories

Although the majority of WebC-Docs’ usage scenar-

ios will likely be based on the server’s local ﬁle-

system for document storage, and a local database

server for additional info (e.g., metadata), WebC-

Docs can nevertheless use additional types of repos-

itories for document storage and search; this can

be useful for scenarios in which an external docu-

ment source already exists (e.g., a DSpace repos-

itory (DSpace, 2009) in an academic context) and

that source repository should be used by WebC-

Docs (instead of using just a local server reposi-

tory). All of WebC-Docs’ interactions with any repos-

itory (whether local or remote) are done through ei-

ther Base Folder Repository or Base Document

Access objects, shown in Figure 8.

The reason why repository access has been di-

vided into these classes, instead of using a single

class with repository access methods, is because all

the information needed to access a document ﬁle, or a

folder repository, is recorded both in the Document

and Folder classes. This allows us to correctly

Figure 8: Support for multiple kinds of repositories.

handle cases in which a document, created in a

folder F1 (using repository R1), is later moved to

another folder F2 (using a different repository, R2).

This avoids moving ﬁles between repositories, which

could present additional issues (e.g., temporary repos-

itory failure). This is also the reason why search

methods are found in the Base Document Access,

instead of Base Folder Repository: to avoid con-

tacting a repository regarding a search for documents

that are not really there.

3 ADDITIONAL ISSUES

AND VALIDATION

Although the previous section has presented some im-

portant aspects of WebC-Docs, it has not presented

some key issues. This section presents a brief discus-

sion of those issues, as well as a few additional notes,

and some case-studies that were used to validate this

system.

A very important aspect in document management

systems (in fact, in any kind of collaborative system)

is traceability. WebC-Docs addresses this aspect in

two complementary ways. Any action performed by

a user (even an anonymous user) is recorded in We-

bComfort’s own logging mechanism, along with all

information that is relevant for that action, so that the

CMS administrator can analyze that log. Also, all

document-related operations are recorded in WebC-

Docs’ own historic records (which are not modiﬁable,

and can be viewed in the Document’s details by any-

one with the required permissions).

WebC-Docs - A CMS-based Document Management System

Document versioning is another aspect addressed

by WebC-Docs (i.e., it is possible to revert a Docu-

ment back to the state in which it was at a certain

point in time), but this is not described here due to

space limitations.

WebC-Docs supports the speciﬁcation of a Doc-

ument’s Authors and Editors through the functional-

ity provided by another WebComfort toolkit, “WebC-

Entities”, that provides concepts such as Entity,

Person, and Group. This allows WebC-Docs users

to specify that “Person X is one of the authors of this

document”, instead of just supplying a meaningless

string of characters (which, in practice, usually do not

really convey much useful information).

Bootstrapping is a typical problem in any sys-

tem that involves conﬁguring permissions; in WebC-

Docs, by default no role has permissions to any Doc-

ument or Folder. The bootstrapping problem here is

in ensuring that, although no particular role has ac-

cess to a Document or Folder (by default), it is still

possible to conﬁgure Folders and Documents. This

has been handled by making the CMS Administra-

tor role (which always exists, and is granted only to

a few select users) be automatically granted access to

any Document or Folder; thus, it is the CMS Adminis-

trator’s responsibility to conﬁgure initial settings and

permissions for WebC-Docs.

Regarding the Dynamic Attributes mechanism

(shown in Figure 5), we consider the following notes

relevant: (1) regarding the “Attribute template” sec-

tion, it is not unlike the relations that the UML meta-

model itself establishes between Class, Property, and

Type (OMG, 2005); and (2) regarding the “Attribute

instance” section, it is not a case of linguistic instan-

tiation, but rather ontological instantiation (which is

explained in (Atkinson and K

uhne, 2003)).

It is important to note that WebC-Docs has already

been validated in a number of case studies, such as:

• SIQuant’s institutional web-site (at

http://www.siquant.pt), as well as (WebCom-

fortOrg, 2009), which were already using the

WebComfort CMS platform, are now using the

WebC-Docs toolkit to make available a variety of

documents (e.g., white-papers, user manuals);

• Our own research group’s document repository

(http://isg.inesc-id.pt/gsidocs) uses WebC-Docs

to make our scientiﬁc production publicly avail-

able to the community;

• WebC-Docs is also being used by a Portuguese

real-estate-related company, to organize and cata-

log their vast paper-based documentation archive.

Finally, it should be mentioned that, although

WebC-Docs has not yet been integrated with other

systems (e.g., DSpace (DSpace, 2009)), it already

allows external sources to access search results by

means of RSS (Really Simple Syndication): a WebC-

Docs search module can make its results available as

a RSS feed, which can be consumed by an additional

(light-weight) WebComfort toolkit called “WebC-

Docs Viewer”, or by any RSS feed reader (even a

regular reader, such as those provided by most cur-

rent browsers, can view those feeds). Document de-

tails are made available in the feed by using the RSS

extension mechanism, and the generated feeds are to-

tally compliant to the RSS speciﬁcation.

4 RELATED WORK

Besides WebC-Docs, other document management

systems exist which handle some of the issues pre-

sented in this paper. In this section, we present some

of which we consider most relevant, while making a

comparison between those systems and WebC-Docs.

OpenDocMan (OpenDocMan, 2009) is a free,

open-source, document management system that is

designed to comply with the ISO 17025 and OIE

standard for document management (OpenDocMan,

2009). Like WebC-Docs, it features a fully-web-

based access mechanism (through a regular web

browser), a ﬁne-grained access control to ﬁles, and

automated install and upgrades. OpenDocMan sup-

ports the concept of “transaction” because any doc-

ument is in either the “checked-in” or “checked-out”

states; WebC-Docs does not use this philosophy on

purpose, because it would make it impossible for

users to even view a document if it was checked out by

a different user (e.g., because the user forgot to check

the document back in). Also, like in WebC-Docs,

OpenDocMan allows its administrator to add addi-

tional ﬁelds besides “document category” to further

describe the document; however, those OpenDocMan

ﬁelds are only strings, while WebC-Docs supports ad-

ditional ﬁelds of any type (e.g., maps) through its Dy-

namic Attributes mechanism.

DSpace (DSpace, 2009) is an “open-source solu-

tion for accessing, managing, and preserving schol-

arly works” (DSpace, 2009). It was designed to be

as standards-compliant as possible (e.g., it supports

the Open Archives Initiative’s Protocol for Meta-

data Harvesting (OAI-PMH) v2.0, and its metadata

schema currently supports the Dublin Core speciﬁ-

cation, although its developers hope to support an

IMS/SCORM subset in the near future).

DSpace supports a ﬁxed tree-hierarchy: (1) com-

munities; (2) collections; (3) items; and (4) ﬁles. Al-

though this hierarchy is immutable, the user-interface

KMIS 2009 - International Conference on Knowledge Management and Information Sharing

can be adapted to “mask” some of its aspects (e.g.,

show a community as an aggregation of similar col-

lections). However, this limitation can make DSpace

unsuitable for some (non-academic) cases – although

DSpace’s objective is to support the scholar commu-

nity, and not a wider enterprise-like community.

It supports the addition of metadata ﬁelds to a

resource-submission form, by adding a “name”–“type

of HTML element to use for input” element to a tex-

tual ﬁle. Additionally, like WebC-Docs, its searching

mechanism takes advantage of the available metadata

(provided with each submitted resource) to provide

more accurate search results. It should be noted that,

unlike in WebC-Docs, some kinds of changes to the

DSpace can only be done through the editing of tex-

tual ﬁles by an experienced administrator.

One of DSpace’s primary goals is to handle the

archive and preservation of documents. To this end,

DSpace supports two different kinds of preservation:

(1) bit preservation, which is like the typical ﬁle-

storage mechanism (the document’s bits are always

the same); and (2) functional preservation, in which

the document’s contents are continuously adapted to

the new formats, accompanying the passage of time

(and subsequent evolution of formats). Also, both

contents and metadata can be exported at any time,

using a XML-encoded ﬁle format. DSpace uses the

Handle system (DSpace, 2009) to provide unique

identiﬁers for each managed resource, ensuring that

its document identiﬁers will be valid for a long time.

KnowledgeTree’s document management system

(KnowledgeTree, 2009) is a commercial system fea-

turing functionalities very similar to what can be

found in current CMS systems (e.g., widgets); in fact,

it could even be considered by some as a “docu-

ment management system-oriented CMS”. However,

unlike other web-based document management sys-

tems, KnowledgeTree’s solution uses a rich-client

(Windows-based) application combined with a web-

application (installed on a web-server, and also pro-

viding a web-services communication layer), which

enables a variety of useful operations (e.g., it allows

drag-and-drop operations between local machines and

remote document repositories). This rich-client per-

spective also makes it able to integrate with Microsoft

Ofﬁce applications. On the server-side, it can index

ﬁles of various types, namely: (1) Microsoft Ofﬁce;

(2) PDF; (3) XML; (4) HTML; (5) RTF; and (6) text.

Its underlying concepts are very similar to WebC-

Docs’, namely the concepts of Folder and Document,

which we believe proves that the “Document and

Folder” metaphor used in WebC-Docs is adequate for

this kind of system. It can have forums associated to

documents, to enable discussions about certain doc-

uments (or even other subjects), a feature that could

easily be found in CMS-based systems (forums are

typical examples of functionality offered by a CMS).

Its metadata-handling mechanism is also very

powerful, allowing metadata of various types, which

can be marked as “required”; nevertheless, WebC-

Docs’ Dynamic Attributes mechanism also supports

these features. The system also allows searching over

metadata and over document contents (if they are in-

dexed), like WebC-Docs. An interesting feature of

this system is its concept of document type, which can

be used to specify document categories, with associ-

ated metadata ﬁelds; while WebC-Docs does not sup-

port this kind of concept (as we believe that there can

be many types of document, which are not mutually

exclusive among themselves), it does support specify-

ing the automatic application of an Attribute Set

to new Documents, on a Folder basis. Also, it allows

the usage of tags and tag clouds; WebC-Docs supports

tags, but only at a fairly basic level, because they can

also be easily replaced by using Dynamic Attributes.

The system does address versioning of documents

and metadata, allowing for a document to be reverted

back to a previous point in time. WebC-Docs also

supports document versioning, but only for regular

metadata (e.g., Authors and Editors, comments); Dy-

namic Attributes are not versioned yet.

Like OpenDocMan, the system allows a “check-

in”/“check-out” philosophy. Nevertheless, this fea-

ture can be regarded as irrelevant, considering that the

system supports the speciﬁcation of workﬂows (this

philosophy can easily be implemented by a system

that supports workﬂows), combined with its support

for document permissions.

Finally, we believe that one of WebC-Docs’ great-

est advantages over KnowledgeTree’s system is its ex-

plicit usage of a CMS extensible platform (WebCom-

fort): functionalities added to WebComfort (such as

an additional toolkit) can easily be integrated with

WebC-Docs (e.g., using forums), while “built-from-

scratch” systems must have such functionality cre-

ated/adapted to their speciﬁc platform.

5 CONCLUSIONS

The expansion of the Internet has originated many

CMS and ECM systems that aim to facilitate the man-

agement and publication of digital contents. How-

ever, most CMS platforms still do not offer function-

ality of a more complex nature (such as document

management), while ECM platforms tend to address

such complex functionality but without taking advan-

tage of the possibilities inherent to a CMS platform.

WebC-Docs - A CMS-based Document Management System

In this paper we have presented WebC-Docs, a

document management system for the WebComfort

CMS platform. Besides typical ﬁle storage function-

ality (which can be found in many regular CMS in-

stallations), this system also provides features such

as: (1) specifying customizable metadata for each

document and/or folder; (2) indexing and searching

documents using a powerful search-engine based on

Lucene.NET, or even using additional search-engines

according to the searched folders; and (3) specify-

ing ﬁne-grained permissions for each document and

folder, in a way inspired by typical ACL mechanisms.

For future work, we plan to introduce additional

extensibility and conﬁguration points to the system,

in order to improve its adaptability to different orga-

nizational contexts. Among others, we intend to add

an explicit conﬁguration point to the system regard-

ing the document indexers that are installed/used in

the system; this will allow us to explicitly conﬁgure

which indexers to use for each installation.

Another aspect to address is the integration with

other document management systems (e.g., DSpace).

Although WebC-Docs supports the usage of addi-

tional data-stores (on a folder basis), connectors to

such data-stores have not yet been developed (only

for local data-stores, such as the local ﬁle-system and

a local relational database). Also, we intend to add

a web-services-based interface to WebC-Docs, allow-

ing other systems to interact with it without explicitly

requiring the usage of its web-based interface.

Finally, another important topic is the speciﬁca-

tion of document management workﬂows. Currently,

WebC-Docs only allows the management of Docu-

ments and Folders, according to the current user’s per-

missions. It would be desirable to specify the various

steps of a workﬂow, in order to adapt WebC-Docs to

more complex application scenarios.

ACKNOWLEDGEMENTS

The authors would like to thank the members of EDM

(Empresa de Desenvolvimento Mineiro, S.A.), as well

as all the other testers of WebC-Docs, for all their hard

work, both in assisting with the testing of WebC-Docs

in a real organizational environment, and in providing

very helpful suggestions for the system.

REFERENCES

AIIM (2009). Association for information and image man-

agement. 1732009, http://www.aiim.org.

Atkinson, C. and K

uhne, T. (2003). Model-Driven

Development: A Metamodeling Founda-

tion. IEEE Software, 20(5):36–41. 2132009,

http://doi.ieeecomputersociety.org/10.1109/MS.2003.

1231149.

Boiko, B. (2001). Content Management Bible. John Wiley

& Sons, Hoboken, New Jersey, U.S.A.

Carmo, J. L. V. d. (2006). Web Content Management

Systems: Experiences and Evaluations with the We-

bComfort Framework. Master’s thesis, Instituto Su-

perior T

ecnico, Portugal.

CSTools (2009). Malcolm Crowe’s Home Page (CSTools).

2132009, http://cis.paisley.ac.uk/crow-ci0/.

DSpace (2009). DSpace.org. 2232009

http:// www.dspace.org.

Kampffmeyer, U. (2006). ECM – Enterprise Con-

tent Management. 1732009, http://www.project-

consult.net/Files/ECM White Paper kff 2006.pdf.

KnowledgeTree (2009). KnowledgeTree Doc-

ument Management System. 2232009,

http:// www.knowledgetree.com.

LuceneNet (2009). Lucene.Net. 2132009

http://incubator.apache.org/lucene.net/.

OMG (2005). Object Management Group – Uniﬁed

Modeling Language: Superstructure – Speciﬁca-

tion Version 2.0. 2132009, http://www.omg.org/cgi-

bin/apps/doc?formal/05-07-04.pdf.

OpenDocMan (2009). OpenDocMan – Free Doc-

ument Management Software DMS. 2232009

http://www.opendocman.com.

Rockley, A. (2002). Managing Enterprise Content: A Uni-

ﬁed Content Strategy (VOICES). New Riders Press.

Saraiva, J. d. S. and Silva, A. R. d. (2008). The WebComfort

Framework: An Extensible Platform for the Develop-

ment of Web Applications. In IEEE Computer So-

ciety, editor, Proceedings of the 34th EUROMICRO

Conference on Software Engineering and Advanced

Applications (EUROMICRO 2008), pages 19–26.

WebComfortOrg (2009). WebComfort.org. 862009

http://www.webcomfort.org.

KMIS 2009 - International Conference on Knowledge Management and Information Sharing