Generating knowledge through Workflow Management Systems



Generating knowledge through Workflow Management Systems

PARASKEVAS J. DALIANIS

Department of Technological Education and Digital Systems

University of Piraeus

M. Karaoli & Dimitriou 80, 185 34, Piraeus

GREECE

Abstract: A systematic approach to the functional requirements of complex Workflow Management Systems is presented in this paper. Their interference with the other inter-organizational information management systems, as well as their implications and consequences are briefly described. An in depth description of the extensions of such systems in terms of Knowledge Management methodologies, required to be implemented in a modern knowledge production organization, is also provided. The extensions required for the efficient extraction and management of knowledge about the organization and operation of the enterprises are presented.

Our approach and the underlying methodology is analyzed through its systematic adaptation in the everyday operations of a national insurance organization located in Greece. The experience gained from this new deployment of the system, provides more results, which proofs the importance of related systems for the improvement of the quality and effectiveness of the services provided by the organization. The full integration of such systems within the electronic working environment is a valuable tool for the improvement of the management infrastructure.

Specific directions towards further involvement of such workflow systems for the automated production of knowledge within a modern enterprise environment are finally proposed.

Keywords: Workflow systems, knowledge management, knowledge processing, business process automation.

Introduction

In the current competitive business environment, organizations need to know what they know and be able to leverage on their knowledge base to surpass their competitors. In this knowledge era, organizations can create and sustain competitive advantage through initiation of appropriate knowledge management processes [13] and leveraging technology [15]. The recent emphasis on knowledge management arises out of the need for organizations to manage resources more effectively in a hyper-competitive, global economy. The need for emphasis on knowledge management is also stressed by Nonaka [7] in the statement “In an economy, where the only certainty is uncertainty, the one sure source of lasting competitive advantage is knowledge. Successful companies are those that consistently create new knowledge, disseminate it widely throughout the organization, and quickly embody it in new technologies and products”. Drucker in [4] declares that knowledge is just not another resource like labor or capital, but is the only important resource today.

In this sense, Knowledge Management has to benefit from any input coming from any processes in an organization that might contain knowledge and information, like material processes, information processes, and business processes Georgakopoulos [5].

Workflow is a concept closely related to reengineering and automating business and information processes in an organization. A workflow may describe business process tasks at a conceptual level necessary for understanding, evaluating, and redesigning the business process. On the other hand, workflows may capture information process tasks at a level that describes the process requirements for information system functionality and human skills. The distinction between these workflow perspectives is not always made, and sometimes the term workflow is used to describe either, or both, of the business and information systems perspectives.

Workflow Management systems (WFMS) are indeed considered as Information Systems (IS), since they are a formal set of business processes, operating on a collection of structured data and contributing part of the information needed for business control and management activities. Furthermore, such systems, partially support decision-making activities and share various features [12], listed as follows:

1. They maintain a database with each instance of the processes to be managed (structured data).

2. The same stages like those for the development of a traditional system have to be performed in order to implement a Workflow solution: Design, Construction and Implementation, following the order established by the development methodology used.

3. In the development stage the same problems are faced: Software Stability, User Resistance To Change, Differences Between The Specifications And What Can Really Be Built.

4. It is an efficient technology; however, its effectiveness has not been proven. Considering it is a new technology, it can be a complete failure if steps against resistance to change are not foreseen.

The numerous advantages offered by the WFMS are reflected in the most important factor for any organization: swifter time response, whether it be customer response time, an internal process, execution time and/or the time taken to finish a task. Consequently, costs are reduced and the company’s global performance is significantly increased.

The main feature of the WFMS is the Workflow solution offered to the IS, which works based on pre-existing processes within the organization and which offers links with its database through a sophisticated easy-to-use interface. Additionally, by enabling the co-ordination of tasks, it facilitates the routing of data supplied by the IS, in turn backing decision-making processes.

Workflow management systems are not considered only as applications (that the actor opens to do a task), but also as components of a cooperative information system [11] characterized by a three-faceted architecture: system, organizational and group collaboration facet [3].

Knowledge Extraction from Workflows

Our approach is based on a simple premise: when people have questions that require “know-how,” the system should match a questioner with someone who can answer the question and share that know-how. This should be done in a way that maps to work activities and takes into account business rules, procedures and ad hoc events (like someone being out of the office, for example). At the same time, the communication should be stored, so that, any advice that is of the explicit type can be captured, and any tacit knowledge that is shared can be rated for value.

Matching available experts and resources to needs, and tracking the resulting workflow and information exchange are fundamental business processes in all enterprises. This model matches inquiries to the right expert, facilitates the right interaction process through its action-oriented workflow, captures the questions, answers and behavior, and provides measurement tools for enhancing the question and answer process. It’s a technically sophisticated approach to knowledge-enabling the tacit elements of an organization.

1 Knowledge Representation and Extraction

Static and pre-defined representation of knowledge is particularly suited for knowledge re-use and offers an interesting contrast against the dynamic, affective, and, active representation of knowledge needed for knowledge creation. The premise of the digitized memory of the past as a reliable predictor of the future success is valid for a business environment characterized by routine and structured change. While the digitized logic and databases can facilitate real-time execution of the inter-enterprise information value chains, their efficacy depends upon real-time adaptation of underlying assumptions to continuously account for complex changes in the business environment. Often such changes cannot be recognized or corrected automatically by computerized systems, as they cannot be pre-programmed to detect an unpredictable future. The adaptability of a Knowledge Management System (KMS) is therefore dependent upon its capability of sensing complex patterns of change in business environments and using that information for adapting the digitized logic and databases to guide decision-making, actions, and resulting performance outcomes.

AI and expert systems based KMS can deliver the "right information to the right person at the right time", if it is known in advance what the right information is, who the right person to use or apply that information would be, and, what would be the right time when that specific information would be needed. Detection of non-routine and unstructured change in business environment would still depend upon sense-making capabilities of knowledge workers for correcting the computational logic of the business and the data it processes. A related challenge lies in tapping the tacit knowledge of executives and employees for informing the computational logic embedded in the KMS. It may be possible to gather information about the decision-making logic from human experts if such decisions are based on routine and structured information processing. AI and expert systems related technologies enable complex computation of specific and clearly defined domain expertise areas by compiling inferential logic derived from multiple domain experts. The challenge of 'scanning the human mind and its sense making capabilities' lies in the problem that most individuals may know more than they think they know. This is particularly true about their information processing and decision-making capabilities related to non-routine and unstructured phenomena and to knowledge that spans multiple domains. The meaning making capacity of the human mind facilitates dynamic adaptation of tacit knowledge to new and unfamiliar situations that may not fit previously recognized templates. The same assemblage of data may evoke different responses from different people at different times or in different contexts. Hence, storing explicit static representations of individuals' tacit knowledge in technology databases and computer algorithms may not be a valid surrogate for their dynamic sense making capabilities.

A case study in the insurance sector

The proposed method was applied in a national insurance broker located in Greece. The methodology was improved here based on our previous experience from an international banking organization [2]. It is well known that Knowledge Management holds considerable promise over the financial sector, which as an industry relies largely on intangible products and services.

The method is based on the platform used internally in the organization. The LivelinkTM platform by OpenText [10], the leading collaborative knowledge management application from Open Text, incorporates a number of features, such as, document and knowledge management and collaboration facilities. The Broker selected its completely integrated workflow system along with the fully Web-based environment to collect and manage tacit information.

The underlying workflow mechanism provides a simple to use and manage web-based environment, which can be easily configurable by a Power User. The system provides full control over the structured data (like due dates and deadlines, task performers, duration, etc.) that may be connected to any enterprise workflow procedure. The seamless integrated workflow mechanism has full access to the content managed by the document management functionality. Thus, the system provides the ability to look also for any unstructured information provided or attached next to any workflow step.

On top of the easy-to-use User Interface through the browser, Open Text is offering its own RAD environment to build components and tools directly integrated into the system and connected with its out-of-the-box functionality. With this, it is possible to implement processes that may monitor the workflow mechanism and collect information related to the content circulated within an active workflow. The mechanism’s conclusions can be stored within the system, providing useful information for a possible reengineering of the organization’s processes. Moreover, the conclusions might be useful in generating sources with guidelines and/or directions for the participants. And if we go a bit further, they can even provide the rules for applying dynamic changes to active workflows. With the appropriate knowledge extraction, the system might even decide to take control and redistribute work items automatically as and when required. Such activities currently require human intervention, and are highly skilled, time consuming and thus expensive processes.

An experienced user may currently consult, for example the questions for support coming from the User to the Help Desk and generate a list of similar cases with specific answers to their questions. In such a way, it might be possible to generate a Frequently Asked Questions list, which might either be periodically updated or sent to the User as a “Tip of the Day” or added to the FAQ database for further use.

1 Help Desk Workflow

One of the processes that were recognized as candidates for being automated was the IT Help Desk (maintained by the EDP department). This will be the process that is presented here for reasons of simplicity and confidentiality restrictions.

The IT staff was spending most of their working time (more than 40%) on supporting the internal users in terms of their working environment, their access to the various software, and so on. Moreover, they were dealing with similar requests coming from their external users (since they are acting as insurance brokers), other insurance agents, who were accessing the same environment through the Web. And it was also observed that the same questions were forwarded to the Help Desk (by phone or in person) most of the times.

In order to gain more benefits from the Knowledge Management tool purchased by the Organization and used for the automation of the client’s request for new insurance contracts, it was decided that this specific process might have been implemented with the workflow mechanism. In a very short period, the implementation followed the requirements analysis phase. The methodology chosen was the one, which was successfully applied some months ago in a banking organization.

[pic]

1. Basic Help Desk Flow Chart

The chart, shown in Figure 1, represents the basic flow of information. Three basic steps are presented here. At the first step the user accesses a form, where he/she can fill the information concerning his/her request for help. The task is automatically being forwarded to the appropriate recipient within the EDP, based on the information provided by the contents and the classification of his/her request. The request is then handled by the member of the IT Help Desk, and forwarded to the interested user.

As a result, a new automated workflow process was available through the Intranet to the internal staff. Figure 2 shows the actual Livelink workflow map with which the Help Desk process was automated.

[pic]

2. Implemented Workflow for Help Desk Automation

Following a very short training on the topic, all users were obliged to send any requests or questions they had through the Knowledge Management environment (Livelink). Their request might contain information, like their office number or their agency number, the topic (question or problem) they wanted to report to the Help Desk, the subject area, and so on.

The questions are arriving at the Help Desk and being archived properly into the system. The request is either solved or escalated. Basic reporting, on number of topics per category was very easily available to the EDP staff. The information was evaluated and decisions were taken about offering basic training to everyday tools for specific users in order to overcome the problems they were facing.

However, the whole process after completing any request was manipulated manually by the EDP staff.

To improve the decision-making steps, a knowledge extraction methodology was proposed. It was developed with the Livelink tools and generated both specific statistics and a new input to the Frequently Asked Questions (FAQ) knowledge base, available within the Knowledge management System.

Information would be extracted from the workflow request based on the predefined categories and the contents of the user’s messages. Their earlier proposed solution might be linked to the corresponding Question. Finally, the information will be available on the Intranet hosted on Livelink.

[pic]

3. Help Desk Flow Chart incorporating intelligent matching

Since, the FAQ database is available, the system may automatically answer to questions coming from the users, in cases where the request matches an already examined case. For example, if a user asks for instructions on how to configure his/her Web browser, and such request was already earlier answered by the IT Help Desk, the Workflow mechanism will find the related FAQ topic and return its hyperlink to the user, closing the case.

[pic]

4. Implemented Workflow for the revised Help Desk Automation

The additional mechanism incorporates the efficient call off the integrated Livelink Search Engine in order to match the contents of the request with the appropriate answer automatically (if such exists). If this is not the case, the original flow will be activated.

The whole revised process is presented in Figure 3.

The actual Livelink workflow map that corresponds to the revised process is presented in Figure 4. In this map the integration (XML based with Livelink Search) and implementation step for the dynamic matching of the information submitted with that in FAQ has been added. In case of a successful match the completed answer is send to the user who made the request.

Among the benefits gained from this implementation are the considerable reduction of the time (reduced down to 20% of their working time!) and effort the IT staff spent for Help Desk and user support and the further automation of an important process within the organization.

Conclusion

In the business environment, true intellectual capital comes from a balanced combination of tacit and explicit knowledge. One of the keys to a successful knowledge strategy is a well-developed knowledge infrastructure that includes people and information, readily accessible through the organization’s infrastructure.

Although the concept of automated workflow management can be traced back for more than 25 years, there are still numerous open research issues, ranging from the organizational impact of workflow technology to integration issues in inter-organizational settings and knowledge extraction.

Future developments may include the development and integration of artificial intelligence (AI) technologies to enhance the capabilities of workflow systems. Some techniques are already applied to specific areas of related tools and services, such as personalization and profiling.

It is expected that intelligent assistants will help towards the development of IT-support environments for knowledge workers. Their role will be of significant value within a business process environment [14]. Hybrid intelligent methodologies [1] may be investigated and applied accordingly.

It is expected that the coexistence of various types of workflow systems within one organization will be a common case. Workflow systems used in the Document and Knowledge management environments must now be integrated with those available in the enterprise resource planning systems, in order to further unify the environment and business concepts within an organization. Their seamless integration is one of the great challenges of workflow research. The knowledge extraction, from the information they carry from one step to the other and from the ways people interact with them, will be another challenge.

Our approach apart form its application into different working environments, will be further extended, such that hybrid methodologies will help towards the knowledge capture and dynamic workflow reorganization from the structured and unstructured information carried through a workflow process.

References

1] Dalianis, P.J. (1996). Artificial Intelligence Approaches to Diagnosis and Robotics Control: Neural Networks and Expert Systems, Ph.D. Thesis. N.T.U.A., Athens.

2] Dalianis, P.J. (2002). Workflow systems and Knowledge Management, 15th National Conference of HELORS (EEEE), October 2002, Tripoli, Greece.

3] De Michelis, G., Joachim W., Woo, C., and Eric, Y. (1998). A Three-Faceted View of Information Systems: The Challenge of Change. Communications of the ACM, vol. 41(12), 64-70.

4] Drucker, P.F. (1994). The Theory of Business. Harvard Business Review, Sept.-Oct., 95-104.

5] Georgakopoulos, D. and Hornick, M. (1995). An Overview of Workflow Management: From Process Modeling to Workflow Automation Infrastructure, Distributed and Parallel Databases, Vol. 3, 119-153.

6] Malhotra, Y., (2001). Expert Systems for Knowledge Management: Crossing the Chasm Between Information Processing and Sense Making. Expert Systems With Applications, vol. 20(1), 7-16.

7] Nonaka, I. and Takeuchi, H. (1995). The Knowledge-Creating Company: How Japanese Companies Create the Dynamics of Innovation. Oxford University Press, New York.

8] Nwana, H., (1996). Software agents: an overview, The Knowledge Engineering Review vol.11(3), 205–244.

9] Nwana, H., Ndumu D., Lee, L. and Collis J. (1999). ZEUS: A toolkit for Building Distributed Multi-Agent Systems. Applied Artificial Intelligence Journal, vol. 13(1).

10] OpenText (2002). Livelink Product Specification, White Paper, .

11] Papazoglou, M. P., and Gunter S. (editors.) (1998). Cooperative Information System: Trends & Directions, New York, NY: Academic-Press.

12] Perez, M. and Rojas T., (2000). Evaluation of Workflow-type software products: a case study. Information and Software Technology vol.42, 489–503.

13] Prusak, L., (2001). Where Did Knowledge Management Come From? IBM Systems Journal, vol. 40(4), 1002-1007.

14] Schurr, H., Sttab, S. and Studer R. (1999). Ontology-based Process Support. Proceedings of the AAAI Workshop on Exploring Synergies of Knowledge Management and Case-Based Reasoning, Orlando, FL.

15] Tsai, W., and Ghishal, S. (1999). Social Capital and Value Creation: The Role of Intrafirm Networks. Academy of Management Journal. Vol. 42(4), 464-476.

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download