HICSS 36 Data-driven Methodology to Extending Workflows to

HICSS 36 Data-driven Methodology to Extending Workflows to E-services over the Internet Presenter: Dickson K. W. CHIU Chinese University of HK Co-authors: S. C. Cheung & Sven TILL Hong Kong University of Science and Technology Data-Driven Methodology to Extending Workflow

Agenda n n E-Services Introduction Meta-model for cross-organizational e-service workflows Extending internal workflows to crossorganizational e-services Cross-organizational exception and event support n Implementation Architecture n Conclusion and Future Work Data-Driven Methodology to Extending Workflow 2

The need of e-Services n n n Economy globalization leads to rising demands in e-Services requires extending internal workflows to cross-organizational workflows. Problems: è è How to extend an internal workflow to a crossorganizational e-service? How to handle data exchange effectively and in a unified manner for both regular and exception parts of a business process? Data-Driven Methodology to Extending Workflow 3

A Motivating Example Data-Driven Methodology to Extending Workflow 4

Goal & Approach n Goal: è è n Automate regular interactions Detect & handle exceptions in cross-organizational communication Approach: è è Extend an internal workflow to a cross-organizational e-service supported by a WFMS Enact e-Services by an event-driven mechanism Data-Driven Methodology to Extending Workflow 5

Meta-Model ECA rule: On event if condition then action Data-Driven Methodology to Extending Workflow 6

Data-Driven Methodology n n n Determine the data requirements of a business process. Identify incoming and outgoing external events of required data. Design Web service interface to process these incoming and outgoing external events. n Support authentication in the workflow process. n Provide process monitoring and directory services. Data-Driven Methodology to Extending Workflow 7

Determine Data Requirement n Construct three relations è è è n publish = {(p, x) | p is a party that publish data item x in some task} subscribe = {(p, x) | p is a party that subscribes data item x in some task} depend = {(x, y) | data item x is derived from data item y} Examples: è publish(End User, Quotation request) è depend(System Configuration, Quotation Request) è depend(Customer Information, Quotation Request) Data-Driven Methodology to Extending Workflow 8

Rules for External Events Identification n Notation ext. Evt(p, d): an incoming external event with data d from party p int. Evt(p, d): an outgoing internal event with data d to party p n Rule: Identification of incoming external event publish(p, d) ext. Evt(p, d) e. g. , ext. Evt(End User, Quotation Request) n Rule: Identification of Outgoing Internal Event subscribe(p, d) int. Evt(p, d) e. g. , int. Evt(End User, Quotation Response) Data-Driven Methodology to Extending Workflow 9

Data Requirement Analysis (cont. ) Construct a directed graph G=(V, E) such that V = set of data item, E = depend relation d G=( {d, e, f}, {depend(e, d), depend(f, d)}) n e f Rule: Identification of responses of incoming external Events è è An outgoing internal event int. Evt(p, e) can be a response of an incoming external event ext. Evt(p, d) if e can be subsequently derived from d. This identifies possible ports for Web services. e. g. , ext. Evt(End User, Quotation Request) & int. Evt(End User, Quotation Response) Data-Driven Methodology to Extending Workflow 10

Possible Web Services n Quotation. Service in: Quotation. Request: Customer. Information; System. Configuration; … out: Quotation. Response Quotation … n Order. System in: System. Order Customer. Number; Quotation. Number; … out: Order. Confirmation Order. Status; Invoice; … Data-Driven Methodology to Extending Workflow 12

Process Monitoring and Directory Services n Provides necessary information to carry out inter -organizational processes. è Monitoring services – snapshot of process status è Description of cross-organizational workflow (e. g. BPEL 4 WS, WSFL, eb. XML) è n Balance trust and security – workflow views Provides “personalized” workflow views via directory services, such as UDDI Data-Driven Methodology to Extending Workflow 13

Cross-organizational Exception and asynchronous Events 1. External Event Reception è è è 2. External exception events incl. refinements change. Order. Request change. Delivery. Date, change. Delivery. Location, etc. Events inheritance hierarchy one capturing Web service per class + “catch-all” Web service Sending Events to Business Partners è (a) request regular service (b) supply event feed è (c) request exception handling (d) pass exception status è Transport: Web service invocation, email, ICQ Data-Driven Methodology to Extending Workflow 14

Cross-organizational Exception and asynchronous Events (cont. ) 3. Event Subscription Service è Publish-subscribe paradigm è List all subscribe-able events at a UDDI directory è Subscribing via publisher’s subscription service è Subscription to: v v 4. Regular events (e. g. price updates) for asynchronous alerts Exception events (order cancellation, delivery delayed) automatic exception handling with ECA-rules Event Chaining Automated receiving and sending of events Relay and distribute an event (e. g. Driver update scenario) EU new Driver EU Example: è PV SI Data-Driven Methodology to Extending Workflow EU 15

E-ADOME with Web Services Data-Driven Methodology to Extending TILL Sven Workflow WSe. BT 2002 -16 16

Conclusion n Practical data-driven approach to extend internal workflow to cross-organizational e-services è è n n Supporting the handling of regular activities, exceptions and events Realized with Web Service technology Event-based meta-model of cross-organizational workflows Generally applicable to advanced WFMS or information systems Data-Driven Methodology to Extending Workflow 17

Future Work n Process adaptation for interoperability n Refinement of the workflow view mechanism n Cross-organization security and transaction n n Real-life E-service / E-commerce applications supply / service chains, financial applications, procurement, insurance, CRM Integration of workflows with enterprise document management systems Data-Driven Methodology to Extending Workflow 18

Questions Your Comments and Questions are welcomed. Data-Driven Methodology to Extending Workflow 19

Contacts S. C. Cheung http: //www. cs. ust. hk/~scc/ Dickson K. W. Chiu http: //www. cse. cuhk. edu. hk/~kwchiu/ Sven Till http: //www. cs. ust. hk/~till/ Data-Driven Methodology to Extending Workflow 20