Rivero L.Encyclopedia of database technologies and applications.2006

390

Multimedia Databases

Mariana Hentea

Southwestern Oklahoma State University, USA

INTRODUCTION

The Internet and technologies such as high-capacity storage devices, broadband telecommunications systems, and multimedia development software systems have accelerated the development of new applications based on the use of multimedia database systems. A multimedia database is a repository of different data objects (text, numeric values, Boolean values, dates, graphical images, video clips, and sound files). Examples of applications based on multimedia databases include news and stock market information on demand; movies on demand; home shopping; medical systems; trademark, patent, and copyright databases; geographic information systems (GIS); weather forecasting; Computer Aided Design (CAD) systems; architectural design; fabric and fashion design; interior design; photographic libraries; art gallery and museum management; law enforcement; criminal investigations; military reconnaissance and surveillance; scientific experiments; and educational systems.

The real-time nature and different kinds, including the size of multimedia data, cause problems for the design, implementation, and management of multimedia databases. Multimedia data requires use of specific technologies to reduce the size of the media data so it could be stored within the database. For example, NASA’s Earth Observation System generates a terabyte (1,000 gigabytes) of data a day. This data comprises images recorded from orbiting satellites by video and infrared cameras that are downloaded to earth. The software chosen by the National Library of Australia is the TeraText™Database System, providing a single access point to over 600,000 images from 18 Australian cultural institutions, including libraries, museums, archives, and galleries representing images related to Australia’s cultural heritage from the late 18th century through to the present day. The TeraText™Database System is known for its scalability, flexibility, speed, and sophistication. The increasing importance of multimedia applications and the introduction of SQL3 in 1999 have favored changes to traditional relational databases. These resulted in the implementation of features to store and manipulate large objects within object-relational databases. One example is Oracle9i, which is an objectrelational database management system that offers ca-

pabilities for holding media data. Through the use of media data types, Oracle interMedia software enables the Oracle9i database management system to manage and deliver image, audio, video, and geographical location data in an integrated fashion with other enterprise information.

The creation, storage, and data modeling techniques for multimedia data objects and retrieval of objects are main issues in the development of multimedia databases. Other issues include authenticity, integrity, media metadata, automatic video capture and editing, mobile media applications, scheduling, and quality of multimedia delivery over a network (Bourbakis, 2004; Shu & Yu, 2004). The following section provides a brief overview of the multimedia retrieval methods.

MULTIMEDIA QUERIES

The representation, structure, and retrieval of objects evolved in time from simple conventional data to more complex multimedia data objects such as video and audio. At the beginning, the retrieval of objects was treated as a single data item using queries based on indexing approaches and identification of the attributes that describe an object. However, this method proved to be inefficient. Retrieval of images from a multimedia database is based on similarities. Modern computer technology can be used to extract images of objects from pictures and videos, but the ability to identify these objects semantically is beyond the capabilities of current research. Because computers cannot add semantic interpretation to images, this must be added manually. Visual queries based on recognizing objects on the basis of their color, shape, or texture and other graphical characteristics are possible. Objects can be indexed using feature values and retrieved based on the similarity to the feature values of other objects. Feature values can be derived for an entire media object or just a specific part of its content. Features that can be measured include shape, color, texture, initial position, and direction and speed of motion. Features can be difficult to quantify and their identification may require complicated and lengthy techniques for extraction. The measurement value for a feature may be a representation of the shape or a histogram representing the color

Multimedia Databases

distribution of the object. Several of the methods for the extraction of feature values require use of advanced algorithms based on artificial intelligence (AI) techniques such as data mining (Shih & Wang, 2004; Thuraisingham, 2001), artificial neural networks and fuzzy logic (Tsai, McGarry, & Tait, 2003), and computer vision (Dunckley, 2003). These techniques are explored to enhance the feature extraction and increase the quality of multimedia retrieval and processing.

If the retrieval of images is to be based on keywords, then either the images must be manually indexed, or the objects within the images must be automatically recognized and appropriate keywords added to the index. Whether automatic or manual indexing is used, the descriptive words added by the indexer may or may not conform to the keywords employed by the user. One approach utilizes two different keyword indexing systems, conceptual keyword and scene description keyword, to retrieve images from a database. However, the developed techniques for indexing and retrieval are abundant, but there are no universally accepted techniques for feature extraction, indexing, and retrieval (Deb & Zhang, 2004).

Currently, new methods are emerging and being implemented for retrieving multimedia data objects (e.g., audio and video) based on their content. Research of content-based retrieval of multimedia information started in the early ’80s with investigations into the retrieval of static images, which in turn was based on pattern recognition research of the ’70s. Research into the retrieval of images from video has received more attention during the ’90s. The approaches to aiding the retrieval of visual images from graphics or videos can be classified into the following categories:

•Keyword, in which the content of the images is described by an indexer using keywords or a textual abstract.

•Content-based image retrieval of features that can be automatically extracted from images. The features are selected according to certain criteria, such as color, texture, or shape; by allowing the user to sketch images and then retrieve similar images from the database; by discerning the motion path of an object; and by identifying one object from its position relative to another. There are a number of commercially available systems, such as IBM’s QBIC and Virage products, which are based on the retrieval of images using their non-semantic features.

such as Infoscope and WebSEEk. WebSEEk is a semiautomatic system for retrieving, analyzing, M categorizing, and indexing visual images from the

World Wide Web.

Many algorithms have been focused on contentbased image indexing, but researchers have been investigating an indexing system based on the spatiotemporal characteristics of objects that appear in multimedia applications. Spatial positions are represented by twodimensional coordinates while temporal relationships are represented by a single dimension, time.

Other systems have been developed for automatically indexing video streams received in real time. The source of the video may be, for example, television broadcasts or security cameras. The incoming video data stream is passed to an event detection module that detects scene changes, significant audio changes, camera operations, and the motion of objects in the video. The event detection module uses this information to define the boundaries of what the authors refer to as events. The event boundaries are specific to the application; in broadcast television these may be scene changes or the appearance of captions, and in security surveillance video the beginning of a new event may be signified by a new object entering the scene. The first frame of an event is used as its key frame. The key frame, the event itself (comprising both video and audio), and its time index will be stored for a certain period of time. The key frames provide a visual index for the viewer. The view of the index displayed can be changed to show a more global view, with an index that includes every key frame. Content-based queries are often combined with text and keyword predicates to get powerful retrieval methods for image and multimedia databases.

Hypermedia video links is another system for con- tent-based retrieval of information from a video database. Raw video data is stored in the video storage. Video clips can be extracted from the video object database using video object software. Developers of hypermedia applications can also use the video object manager to retrieve and play video data objects. The video data structure for the hypermedia multimedia authoring environment is quite straightforward. Using the video object manager, the developer of an application can extract clips manually from a video to create primitive video data objects. Attributes, such as display size and frame rate, inherited from the video from which it is extracted can be associated with the clip or altered by the developer. Complex data objects can be created

media management products, examples of applications using these products, and significant standards.

KNOWN MULTIMEDIA MANAGEMENT PRODUCTS

Corporations such as IBM, Virage, Silicon Graphics, ARDA, Oracle, Autonomy, and ALPHATECH are major providers of multimedia database management, communication, and content management software for corporations, media and entertainment companies, universities, and government agencies worldwide.

The IBM’s CueVideo project provides technologies to automatically summarize and index videos and to make them much easier to browse. One approach is using audio stream for search and using video stream for quick browsing in a complementary manner to provide the desired video search functionality (Brown et al., 2001). CueVideo is an ongoing research project to address the challenges of large video databases.

One example of a mobile multimedia application is MobiDENK (Krosche, Baldzer, & Boll, 2004) for monument conservation. The Hermitage museum’s Web site uses an IBM product, QBIC (Query By Image Content), for searching archives of world-famous art. QBIC is a system developed by IBM to explore various contentbased retrieval methods. Queries in QBIC can be based on color and texture patterns, user-drawn sketches, example images, camera and object motion, or other graphical information, such as the color balance; e.g., retrieve all images with 30% blue and 10% yellow colored pixels. The QBIC database can handle both still images and video clips. In common with other systems, videos are divided into shots, using techniques similar to those described above. One frame from each clip is extracted or generated as a representative frame for the clip. In a motion sequence there may not be a single frame that is representative of the entire shot. In this case, QBIC constructs an image of the complete background from the sequence of frames and on to this superimpose images of the foreground objects.

QBIC’s capabilities are available in DB2 Image Extenders, which are components of IBM’s scalable, multimedia, Web-enabled DB2 Universal Database. IBM developed different multimedia retrieval applications for the DB2 Universal Database as follows:

•DB2 Audio Extender enables audio retrieval.

•DB2 Image Extender enables image retrieval. Visual features such as color and texture patterns are used for search criteria. For example, a photographic database may be queried for thumbnail

Multimedia Databases

images of all pictures stored in GIF format, and the name of each picture’s photographer can be listed. Then, Image Extender invokes a browser.

•DB2 Video Extender enables video retrieval. Video and traditional business data may be included in a single query.

For example, Image Extender is used to store print ads, Audio Extender for broadcast ads, and Text Extender for ad scripts. The user can retrieve all the objects in a single query and then preview video ads as video storyboards using the Video Extender.

Hillsborough County uses Virage’s products to create a central digital archive of government proceedings, which automates a formerly manual workflow and provides real-time access to content. Virage products include capabilities as follows:

•IDOL provides the infrastructure for capturing, feeding, and delivering rich media through the enterprise.

•VS Archive is a software solution used by enterprises to store, categorize, manage, retrieve, and distribute video, audio, and other rich media content. It includes new features such as advanced conceptual retrieval and automated hypertext linking to enterprise information found in more than 300 different data types.

•VS News Monitoring is a real-time monitoring and content management solution used by enterprises and government agencies to automatically track content for time-sensitive, strategically significant events. New features within VS News Monitoring include real-time data access and advanced concept-based retrieval. Concept-based retrieval is especially important for monitoring and searching news feeds in foreign languages, because users may work in second languages or rely on transcriptions that may contain misspelled words. With IDOL, VS News Monitoring not only proactively alerts users to broadcast news but also delivers related internal content as well as information from Web sites.

•VS Webcasting is an enterprise software solution that simplifies and streamlines the entire workflow for producing live Webcast events and on-demand or rebroadcast presentations for large audiences.

Silicon Graphics and Virage combine products to create breakthrough media management systems. For example, Virage Media Management System and the Silicon Graphics StudioCentral asset management system enable companies to automatically and intelli-

Multimedia Databases

gently catalog large libraries of videotape and multimedia content into a compact, online database. The combined products provide users with a complete media management system to find and manage their media through a simple Web browser. The next section discusses main standards for multimedia applications.

MAJOR STANDARDS

Many aspects of the multimedia information life cycle are affected by regulatory compliance (Golshani, 2004), and requirements for latest object database management development standards are proposed (Cardenas, Pon, Panayiotis, & Hsiao, 2003). MPEG-7 and MPEG21 standards have been essential for the development of multimedia applications. The MPEG-7 standard, formally named Multimedia Content Description Interface, provides a rich set of standardized tools to describe multimedia content. Both human users and automatic systems that process audiovisual information are within the scope of MPEG-7. MPEG-7 offers a comprehensive set of audiovisual description tools (the metadata elements and their structure and relationships, which are defined by the standard in the form of descriptors and description schemes) to create descriptions (i.e., a set of instantiated description schemes and their corresponding descriptors at the user’s will), which form the basis for applications enabling efficient access (search, filtering, and browsing) to multimedia content (Chang, Kikora, & Puri, 2001).

MPEG-21 has established a work plan for future standardization. Nine parts of standardization within the multimedia framework have already started.

FUTURE TRENDS

Content-based retrieval is a very active area of research despite the advances that have been made. Recently, Virage announced its participation in the Video Analysis and Content Extraction (VACE) media communication and content program sponsored by ARDA. The research project will be focused on the development of video content extraction technology. The initial focus of the VACE program is to develop automatic detection and recognition technologies from various video-re- lated sources, including indoor and outdoor activities, unmanned air vehicles, and television news broadcasts. Over time, VACE technologies aim to provide significant improvement in indexing and retrieval, understanding, image processing, data mining, filtering and selection, and storage and forwarding mechanisms. In addi-

tion, research is occurring on new compression techniques, such as wavelet, vector, and fractal methods, to M ensure multimedia delivery with high quality.

In the future it should be possible to automatically recognize and identify objects that appear in still images and video. To achieve this, it will almost certainly require significant developments in the application of artificial intelligence techniques to multimedia database systems. These developments will lead to automatic recognition and indexing of video footage and still images and result in the development of a wide range of applications relying on the content-based retrieval of multimedia data objects.

CONCLUSION

Multimedia is becoming present everywhere: at home, business, school, hospital, road, etc. Digital media has been acknowledged as a standard data type, allowing for increased personal communications, business-to-em- ployee, business-to-business, and business-to-consumer applications. These applications require complex and large multimedia databases, causing changes in computing and solution architectures to be maintained by organizations. In addition, technologies pertaining to communication, coding, compression, content distribution, storage, mobile computing, media servers, cryptology and watermarking, and digital media management will grow in order to address solutions for multimedia services. While creating digital media is not expensive, it is generally expensive to manage and distribute it. Browsing large multimedia databases can become complex and will demand faster and more efficient algorithms for indexing and retrieval of structures. In addition, multimedia distribution in a mobile computing environment will continue to be the center of research for nextgeneration multimedia systems.

REFERENCES

Bourbakis, N. (2004). Digital multimedia on demand.

IEEE Multimedia, 11(2), 14-15.

Brown, E. W., Srinivan, S., Coolen, A., Ponceleon, D., Cooper, J. W., & Amir, A. (2001). Towards speech as knowledge resource. IBM Systems Journal, 40(4), 9851001.

Cardenas, A. F., Pon, R. K., Panayiotis, A. M., & Hsiao, J.-T. (2003). Image stack stream viewing and access.

Journal of Visual Languages & Computing, 14(5), 421-441.

Chang, S.-F., Sikora, T., & Puri, A. (2001). Overview of the MPEG-7 standard. IEEE Transactions on Circuits and Systems for Video Technology, 11(6), 688-695.

Deb, S., & Zhang, Y. (2004). An overview of contentbased image retrieval techniques. 18th International Conference on Advanced Information Networking and Applications (AINA’04) (Vol., 1, pp. 59-64).

Dunckley, L. (2003). Multimedia databases: An ob- ject-relational approach. London: Addison-Wesley.

Golshani, F. (2004). Multimedia information lifecycle management. IEEE Multimedia, 11(2), 1.

Krosche, J., Baldzer, J., & Boll, S. (2004). MobiDENK— Mobile multimedia in monument conservation. IEEE Multimedia, 11(2), 72-77.

Shih, T., & Wang, P. P. (2004). Intelligent virtual world: Technologies and applications in distributed virtual environment. Hackensack, NJ: World Scientific.

Shu, W., & Yu, M.-Y. (2004). Resource requirements of closed-loop video delivery. IEEE Multimedia, 11(2), 24-37.

Thuraisingham, B. (2001). Managing and mining multimedia databases. Boca Raton, FL: CRC Press.

Tsai, C.-F., McGarry, K., & Tait, J. (2003). Using neurofuzzy techniques based on a two-stage mapping model for concept-based image database indexing. IEEE Fifth International Symposium on Multimedia Software Engineering (ISMSE’03), 1, 6-12.

Almaden (n.d.). Retrieved July 22, 2004, from http:// www.almaden.ibm.com/projects/cuevideo/shtml

IBM (n.d.). Retrieved July 25, 2004, from http:// wwwqbic.almaden.ibm.com

Infoscope (n.d.). Retrieved July 25, 2004, from http:// www.infoscope.com

Multimedia Databases

NASA Earth Observatory, http://earthob servatory. nasa.gov:8000/Laboratory/index.html, accessed July 23, 2004.

Oracle, http://www.oracle.com , accessed July 20, 2004.

TeraText, http://www.teratext.com.au, accessed July 23, 2004.

Virage, http://www.virage.com, accessed July 26, 2004.

WebSeek, http://www.ee.columbia.edu/~sfchang/research, accessed July 22, 2004.

KEY TERMS

Content-Based Retrieval: Method for automatic multimedia content features extraction.

Feature: An attribute derived from transforming the original multimedia object by using an analysis algorithm; a feature is represented by a set of numbers (also called feature vector).

Feature Extraction: Use of one or more transformations of the input features to produce more useful features.

Feature Selection: Process of identifying the most effective subset of the original features.

Indexing: Mechanism for sorting the multimedia data according to the features of interest to users to speed up retrieval of objects.

Metadata: Information about multimedia data objects, applications, processing, and delivery requirements.

Multimedia Database: A repository of different data objects such as text, graphical images, video clips, and audio.

INTRODUCTION

The building of decision support system (DSS) for integrated workgroups sets out problems related to the nature of the decision process as well as its operative implementation.

The multiparticipative decision systems (MDM) abound in the management of companies and public/ private institutions. In these companies, the partial decisions of a group of people in their responsibility field reciprocally influence and determine the operation of the system. The achievement of the strategic goals requires unit of vision and interpretation of those objectives.

The balanced scorecard (BSC) is a tool for unifying the vision and allows a common element to interpret reality, and the environment changes for a heterogeneous group that has to decide among the multiple possible alternatives of a strategic plan.

It is necessary to have at one’s disposal the knowledge and the integration of information and decisions through models of collaboration and integration.

This research work analyzes the usage of BSC as a link element between many decision makers, having different responsibilities and perspectives, who share strategic goals.

This usage requires becoming efficient in an accurate integration of:

a.Collaborative-task environment

b.Shared access to databases

c.Techniques to get new information (relationships, indicators, ratios) such as OLAP, datawarehouse, and data mining

d.Decision makers constant learning (from the obtained results)

In these models, it is possible to play down the interpretative divergence of the personal differences of the decision makers.

The analysis of the applications, decision models, integration and use of these systems computer-supported cooperative work (CSCW) constitutes a prosperous field to develop management computerized structures.

To integrate the functions of a company into a unified strategy, work models facilitate the spread of knowledge. BSC is much more than an information system; it is a knowledge management system which allows adapting, in a coordinate way, those decisions leading to the achievement of strategic goals. Thus, integrated techniques, tools, and working methodologies will also make possible:

a.A rise in shared models

b.A rise in learning speed

c.Improvement in the quality of decisions

d.Major action coordination

In these models, it is possible to play down the interpretative divergence of the personal differences of the decision makers.

BACKGROUND

Multiparticipant Decision Making

We can define it as “an activity conducted by a collective entity composed of two or more individuals and characterized in terms of both the properties of the collective entity and of its individual members”. Each of the MDM types has a specific structure of interaction among the decision makers and the participants (Marakas, 1997)

The MDM is a common and growing reality. Diverse modern business structures such as virtual companies, the strategic alliances, the “cooperation-competence” relations, and the new philosophies and management modes require forms, methodologies, and techniques applied to the decisions that surpass space and time limitations.

This growing development of management concepts significantly collaborates informatics. The aforementioned limitations imply others which are the ones that harm the quality, swiftness, and comprehension of decisions such as:

a.Shared mission

b.Unified vision

c.Availability of information

d.Criteria unification

e.Acceptation of particular aims

f.Coordination

g.Feedback (awareness of the effects and the decisions)

The different types and classes of support technologies show the different alternatives and possibilities of improving quantitative tools and informatics. Those decisions should be shared by many people.

BSC offers the possibility of joining them according to their application opportunities without detriment to the utilities of those techniques and tools, in a causeeffect analysis relative to the tangible and intangible components of the strategic decisions.

Figure 1. Scheme of decision processes

Scheme of Decision Processes

Simon (1960) elaborated a model about the problemsolving characteristics, which represents a permanent scheme of the decision-maker’s activity (Figure 1).

Decisions Based on Strategy

These decisions are of particular interest in our case since we can consider the strategy as Byrnes and Chesterton (1978) and Mintzberg (1993) did:

a.Plan

b.Action rules

c.Behavior pattern

d.Position

e.Perspective

Since these are strategies useful to make decisions on tactic plans and to establish actions consistent with those ones, it is of particular importance that their comprehension, alignment, and participation of those who are responsible for carrying it out (Davenport & Prusak, 1997; Steiner, 1997).

Comprehension and involvement need an “integrated knowledge environment” to elaborate and apply the chosen strategies. (Tissen, Andriessen & Lakanne, 2000)

Factors, Conditions, and Perspectives

Affecting Decisions in Organizations

Factors

Decisions in an organization are affected by factors characteristic of the relationships established within the organization and with its environment, such as in Marakas (1999) and Holsapple and Joshi (2001):

a.Group structure

b.Roles

c.Processes

d.Decision styles

e.Norm and decision rules

These characteristics have non-quantitative main components; then it is necessary to get a hold of the right conceptions for their coordination and treatment, orientated to optimize team work decisions and to improve those relationships which help to encourage group synergy (Fleischer & Mahaffy, 1997).

Group Decision-Making Conditions

Group decisions are taken under conditions of limited knowledge, uncertainty, poor communication, option unawareness, and diverse goals (and sometimes contradictory) of the decision makers.

The difficulties which generate the conditions previously mentioned tend to increase considering the different levels (operative, tactic, and strategic) in which they have to be implemented (DeSanctis & Gallupe, 1987).

Modern knowledge management systems are supposed to meet different decision makers abilities as they face common problems and shared decisions. BSC comes out as a model gathering different techniques, perceptions, and approaches through perspectives shared by all the group members in strategic planning adaptable to the changing conditions of the environment and the organization itself (Pigott, 2000). These models represent a sort of “virtual community”, qualified by knowledge of the reality towards a specific strategic goal (Tissen et al., 2000).

Decision Perspectives

Keen and Morton (1978) have stated a classification of the perspectives of a decision-making process which shows the extent and the transcendence of its improvement:

a.Administrative rationale perspective

b.Processes orientation perspective

c.Organizational procedures perspective

d.Political perspective

e.Individual difference perspectives

All of the decision perspectives have the characteristic of information and knowledge.

As information because they are a series of formal rules such as descriptions of processing, probabilities of occurrence, and representations of meaning by means of symbols or mathematic operations.

As knowledge because they are related with description, analysis, comprehension of the environment, and any probable change necessary for organization strategies.

Knowledge in organization bears the following characteristics (Barnes, 2002):

a.Comprehension and reflection

b.Communication and broadcasting

c.Knowledge aspects and results

CRITICAL ISSUES AND PROBLEMS

Decisions and Knowledge

As knowledge has its source in learning, it broadens considering three main components:

a.Cognitive learning

b.Social learning

c.Feedback learning

As an exponential factor of decisions intelligibility (I3) (Tissen et al., 2000).

Each learning component has an important part of shared knowledge, the three of them a consequence of interaction and mutual reinforcement in understanding reality and its problems.

Basing in these concepts to generate (make) shared decisions means a change in the culture and attitudes which require self/own elements to materialize in conducts and results (Neidorf, 2002).

Knowledge Management Systems

M

A knowledge management system (KM) should be useful to whoever is implicated in the comprehension processes, evaluation and reorganization of the company. (Frank, 2002)

This means offering relevant knowledge to groups conformed by executives, analysts, technicians and assessors, employees, customers, and suppliers who take part in the processes.

Knowledge management systems must fulfill the following requisites (Frank, 2002):

a.Conceptual level emphasis

b.Existing knowledge reuse

c.Individual needs adaptation

d.Intuitive comprehension

e.Different perspectives

f.Perspectives integration

One of the main challenges of the knowledge management systems is the transformation of implicit knowledge (subjective, hard to recognize and transmit) in explicit (one/knowledge) and its transference and integration (Nonaka & Takeuchi, 1995).

Beyond conceptual requirements, KMs must have technical features which make them useful to decision groups, such as in Alavi and Leiner (1999), KPMG (1998), and Easterby-Smith (1997):

a.Information capable of turning into actions

b.Categorized data for usage

c.Information filters

d.Information accessibility

e.Data storage systems (data warehouse)

f.Integrated databases

g.Extraction information techniques (data mining)

h.Analytical decision techniques

i.Facilitate constant learning

j.Facilitate group and organizational learning

k.Advance information and communication technologies

BSC as Knowledge Management

System

The possibility of getting the previously mentioned learning modes greatly depends on the utilization techniques and tools which enable getting hold of datagenerated knowledge and make use of it from an interpretative level or state of reality.

Balanced Scorecard Collaborative (BSC) is basically a strategic management system which allows by means of cause-effect relationships and the indicators which represent them (Kaplan & Norton, 1997):

a.Clarified and translate or transform vision and strategy

b.Communicate and link strategic objectives and indicators

c.Plan, establish objectives, and align strategic initiatives

d.Increase feedback and strategic formation

For each perspective, the executors have quantification elements of the relevant variables and the possibility of deepening their analysis by means specific applications, having at the same time a global view of the strategy and the degree of general realization.

Some important aspects in BSC as knowledge management systems are those of principal utility in the elaboration of group decisions.

The data and information expressed may be cleverly interpreted and used for the decision by different people from different points of view:

a.Its offers unified vision

b.It facilitates knowledge interchanging

c.Promotes learning in different levels and ways

Informatic Requirements in Applying

BSC in MDM

Virtual Collaborative Environment

The overcoming of these limitations is especially favored by the implementation of Virtual Collaborative Environments (EVC) where the application of management technologies plays a vital role.

Groupware and workflow techniques belong to that framework.

The combination of these technologies with strategic management tools, such as BSC, can provide the shared knowledge conditions, feedback, and adapted to the approaches heterogeneity, capacities, and interests of the integrants of the decision-making group.

Network development (Internet, Intranet, Extranet) opened new possibilities to organization management and informatics usage. The tools previously mentioned, fundamental in collaborative task environment, are included in this category.

Business Process Component Model (BPCM) shows the link reality-information-evaluation-decisions, incorporating a BSC for the analysis (Figure 2).

Figure 2. BSC in BPCM

Evaluation

Decisions

Reality

The model contains

1.Communication

2.Collaboration

3.Coordination

which are requirements of a system of decision generating in the business processes. The model will require a group of interfaces and formats to define its operation from a chosen strategy.

Shared Strategy Control Model

Different BSCs can be applied to the integrant parts of a general strategy which joins and controls them in a shared BSC (Figure 3) (Kaplan & Norton, 2001).

Group Decision Support Systems (GDSS)

They are interactive systems based on computers to sort out non-structured problems by a group of decision makers.

A GDSS provides substantial improvements in the operation of decision groups such as in Landon and Landon (1996):

a.Increase the precision, knowledge, and effectiveness of the planning

Figure 3. Strategy and BSC

BSC

Reality

b.Increase the participation and the integration

c.It is a collaboration tool

d.It is useful to generate ideas

e.Increase the objectivity of the evaluation

f.Establish priorities

g.Improve the organization

h.Generate and increase the availability of necessary documentation

i.Facilitate access to external information

j.Promote institutional development

k.Reconcile the concepts and information used by the integrants

Apart from the great elaboration and complexity models, among the most used in GDSS are:

a.Electronic questionnaires

b.Benchmarking tools

c.Organization of ideas

d.Vote techniques or priorities

e.Leader identification and analysis tools

f.Policy formulation tools

g.Group dictionaries

h.Electronic meeting systems

As it can be appreciated, the possibilities of achieving an effective system of interaction leading to the best decisions which are wide and varied.

GDSS in Network Structures

Participative models using BSC can also be used in business networks. A clear example is the linked tourist services (transportation, lodging, shows, excursions, etc.).

These networks need appropriate information processing systems for fluidity, flexibility, and originality of the decisions (Serra & Kastika, 1994). The dynamics of information, the cooperation for achieving results, and the permanent adaptability to the market requirements need tight links of knowledge and management.

A model with an adaptable structure to the requirements and the development of competence cultural guidelines which allows to establish global shared strategies can be set up over these networks.

In these common strategies, the specific ones which belong to each participant in the network will be inserted.

Binding Scheme of the Group Decision

Support System (GDSS)

A possible architecture for multiparticipative decisions through knowledge management (Figure 4) shows the

BSC

characteristics of the system which has the following properties (Abecker et al., 2002):

a.Active help in the search and detection of information

b.Integration functions (data, formal knowledge, representation, and knowledge materialized in products and processes)

c.Self-adaptation and self-organization

BSC Advantages and MDM Use

To the already mentioned characteristics of BSC and GDSS as specific of the applications, the advantages originating in the functional usefulness of their combination should be added.

In a non-exhaustive enumeration, the following points could be mentioned:

1.Regular monitoring of the results of the strategy expressed on the BSC.

2.Immediate knowledge of the consequences and individual-group actions (Malina & Selto, 2001).

3.Detection through the established indicators of the difficulties inconsistencies, and diversions, and other flaws of the chosen strategy (Kaplan & Norton, 1997).

4.Possibility of detecting, through the application of Data mining and data warehouse strategies, new indicators that evidence unconsidered relations the moment of setting out a strategy. The development analysis and individual applications through a shared database in which each participant can evaluate decision field and integration in the context of development of the strategy.

5.It allows evaluation of intangibles with conceptually homogeneous criteria, incorporating them as a principal factor in strategic management. This idea