Business Capability-centric Management of Services and Process Models (Extended Abstract) Wassim Derguech Insight Centre for Data Analytics - National University of Ireland, Galway wassim.derguech@insight-centre.org Abstract. This research details a concept model for describing the busi- ness capability of services and business processes from a functional per- spective in terms of actions together with related business properties. Furthermore, this work proposes the aggregation, indexing, discovery and design of configurable models for services and business processes using the concept of business capability. Keywords: Business Capability, Services, Business Process, Aggrega- tion, Indexing, Discovery, Configuration 1 Introduction With the advent of Industry 4.0, more and more companies are actively work- ing on digitising their assets (i.e., services, processes, etc.) for better control, collaboration, modularity, analysis, etc. By 2020 more than 80% of companies will have digitised their business processes and value chains. This creates more services and processes, making their indexing, discovery, configuration, etc. more challenging. Thus, properly digitising those assets needs a proper data model to describe them, and proper algorithms for indexing, discovery and configuration. In this context, this research proposes a concept model for describing the business capability of services and business processes from a functional perspec- tive in terms of what do they achieve together with related business properties [7, 1]. Furthermore, this work proposes the aggregation [6], indexing and discov- ery [8], and design of configurable models [5] for services and business processes using the concept of business capability. The remainder of this paper is organised as follows: Section 2 introduces the business capability meta-model used for designing domain specific capabilities for the annotation of services and process tasks. Section 3 gives on overview of the capability aggregation algorithm proposed as part of this work. Section 4 investigates the use of Formal Concept Analysis for the indexing and discovery of business capabilities. Section 5 discussed the use of business capabilities in the design of configurable process models. Finally, Section 6 concludes the paper and discusses future research directions. 2 Towards a Structured Business Capability Description The first objective of this work is to propose a conceptual model for describing Business Capabilities [7, 1]. The proposed model is implemented as a set of on- tologies that can be used for creating semantic annotations of business process models or services. In this work, I consider a business capability as standalone entity that can exist outside the scope of service descriptions or invocation in- terfaces. A service, a computer program, a business process or even a manual task can be described using the business capability concept. In a very simple definition, I consider a business capability as a set of actions enriched with zero or many properties. Properties allow to refine further the action that is taken for a domain related ontology. More formally, as shown in Fig. 1, in the proposed model capabilities are defined as an Action Category and a set of property entries. The Action Category concept is similar to an action verb [14] that defines, in a natural language, what is the action being described. Different to the concept used by Oaks et al. [14], I consider the category as a concept from a domain related ontology that comes form a shared agreement on its semantics. A category is a specific property that is present in all busi- ness capability descriptions via the property achieves (See Fig. 1). A property entry, defined with respect to certain declaration, is a couple (property, value) where property is a domain-specific functional feature or a domain-independent non-functional property and value is the value or the possible values that a prop- erty can have. Both property and value refer to ontological terms. For example, shipping services can be described using the action category “shipping” that can be extended with properties reporting on the “source address”, “destination address” , etc. 1..* achieves Action Category Capability specifies extends properties 0..* definedW.R.T Property value 1 Property Entry Value Declaration 1 0..* PropertyName : String PropertyName : String SpecRelation : Set Configurable: Boolean Fig. 1. Business Capability UML Class Diagram The model was validated via Bunge’s theory of ontology [2] to verify its syntactic correctness: no constructs’ redundancy and no semantic ambiguity. Interviews with domain experts show that the model is suitable for end-users understanding. However, tool supports are required for facilitating its adoption. 3 Aggregation of Business Capabilities: Determining the Business Capability of a Process Model The second objective of this research is to propose an abstraction technique that allows moving from an entire process model to its functional description by aggregating the business capabilities of the process elements into a single one. This feature is necessary when a user wants to have an overview of the capability of an entire business process model rather than a single activity. In this context, the contribution is an algorithm that automatically gener- ates the aggregated capability of an entire capability-annotated business process model [6]. It is a capability propagation algorithm that starts from the initial node of a process model and traverses the model by firing one node at a time. Each node introduced new changes to the propagated capability. The algorithm is formally verified with formal semantics using Petri Nets [15] and implemented in a tool to support the validation by interviewing domain experts. The experts find that the approach is very well aligned with their expectations, the results of the aggregations are easy to interpret/understand and the tool support was simple and intuitive to use. 4 Indexing and Discovering Business Capabilities The third objective of this work is to explore the use of Formal Concept Analysis [11] (FCA for short) for providing efficient indexing and discovery of business ca- pabilities that are described using the proposed model [8]. Rather than inventing a new set of indexing and discovery algorithms, I reused FCA as a mathematical classification tool. Using FCA has the advantage of benefiting from already well established indexing and search algorithms that I have adopted for discovering a set of service descriptions (i.e., capabilities). In the evaluation of this work [8], I used in a first experiment, a set of real- world sensor services descriptions to validate its applicability in managing a reduced number of capabilities. In a second experiment, I used synthetic data set and measured the time required to create and traverse an index of a large number of capabilities with different configurations. Results of the evaluations show that the approach is effective and performs better than related approaches as shown in Table 1. 5 Using Business Capabilities in the Design of Configurable Business Process Models The fourth objective of this research is to reduce the business process modelling effort when using configurable process models by proposing an algorithm for creating business capability-annotated configurable business process models that capture configuration options in terms of business capability features. After the analysis of reuse-oriented business process modelling techniques in order to find how capabilities were used and how can the business capability Table 1. Comparing Time Performances of Indexing Approaches Indexing Mechanism Time Performance Inheritance between OWL-S services [3] N/A Topic extraction and Formal Concept size: 1088 services, query response time between 300 Analysis [13] and 3000 ms size: 50 services, index construction + advertisement Reasoning-based matchmaking [18] time: ∼ 4 s Numerical encoding of ontological con- size: 100 services, index construction + advertisement cepts and codes comparison [17] time: ∼ 500 ms Capabilities Indexing using Formal Con- size: 1000 capabilities, index construction + parsing cept Analysis [8] time: ≤ 25 ms model contribute to this area [9], I found that configuration-based modelling was a suitable starting point. The idea is to start from a reference process model [16] (called a configurable model) and tailor this model to meet the end-user needs by enabling or disabling several branches of the model. The current state of configurable model requires an extensive process mod- elling techniques to carry out proper configurations (captured in terms of model restrictions and parameters) or extensive manual effort for creating an abstract configuration layer [12] that reflect the business needs of the end-user. I suggest in this work the early integration of business capability descriptions of activities in process models and proposed an algorithm for creating config- urable models that capture configuration options from a functional perspective (capabilities parameters) in order to shift the configuration from manipulating the model directly to manipulating the parameters of its capabilities. My pro- posed solution [5] has been tested on real world business process models from municipalities (used in existing case study [10]) and customs clearance proce- dures. Furthermore, I carried semi-structured interviews with domain experts to assess the usefulness of the proposed approach. Results show that the experts are familiar with reference models and the proposed work simplifies the understand- ing of configuration options and the impact of configuration decisions. However, the current state is not mature enough to be integrated in their systems. 6 Conclusion The core contributions of this work can be summarised as follows: (1) a capabil- ity conceptual model for describing the action performed by services and process models, (ii) an algorithm for the aggregation of business capabilities that is com- parable to process abstraction techniques, (iii) the validation of the applicability of formal concept analysis for the efficient indexing and discovery of capabilities and (iv) an algorithm for automatically creating a capability-annotated con- figurable process model by merging a set of process variants. Further research directions can be explored for each of these contributions that have beend dis- cussed in details in my thesis [4]. Bibliography [1] Sami Bhiri, Wassim Derguech, and Maciej Zaremba. Modelling Capabilities as Attribute-Featured Entities. In WEBIST 2012, LNBIP. Springer, 2012. [2] M. Bunge. Treatise on Basic Philosophy. Ontology I: The Furniture of the World. Boston, Riedel, 1977. [3] Daniel Elenius et al. The OWL-S editor - A development tool for semantic web services. In ESWC 2005, LNCS. Springer, 2005. [4] Wassim Derguech. Business Capability-centric Management of Services and Process Models. PhD thesis, NUI Galway, Ireland, Dec 2016. [5] Wassim Derguech and Sami Bhiri. An Automation Support for Creating Configurable Process Models. In WISE 2011, LNCS. Springer, 2011. [6] Wassim Derguech and Sami Bhiri. Business Process Model Overview: De- termining the Capability of a Process Model Using Ontologies. In BIS 2013, LNBIP. Springer, 2013. [7] Wassim Derguech and Sami Bhiri. Modelling, Interlinking and Discovering Capabilities. In IEEE AICCSA, 2013. [8] Wassim Derguech, Sami Bhiri, Souleiman Hasan, and Edward Curry. Using Formal Concept Analysis for Organizing and Discovering Sensor Capabili- ties. The Computer Journal, 58(3):356–367, 2015. [9] Wassim Derguech, Sami Bhiri, and Edward Curry. Techniques for reuse in business process modeling in public administration. In Government 3.0 - Next Generation Government. 2017 (to appear). [10] Florian Gottschalk et al. Configurable process models: Experiences from a municipality case study. In CAiSE 2009, LNCS. Springer, 2009. doi: 10.1007/978-3-642-02144-2 38. [11] Bernhard Ganter and Rudolf Wille. Formal concept analysis - mathematical foundations. Springer, 1999. [12] Marcello La Rosa. Managing Variability in Process-Aware Information Sys- tems. PhD thesis, Queensland University of Technology, Brisbane, Aus- tralia, April 2009. [13] Mustapha Aznag et al. Leveraging formal concept analysis with topic cor- relation for service clustering and discovery. In IEEE ICWS, 2014. [14] Phillipa Oaks, Arthur H. M. ter Hofstede, and David Edmond. Capabilities: Describing What Services Can Do. In ICSOC, LNCS. Springer, 2003. [15] James Lyle Peterson. Petri Net Theory and the Modeling of Systems. Pren- tice Hall PTR, 1981. [16] Michael Rosemann and Wil M. P. van der Aalst. A configurable reference modelling language. Inf. Syst., 32(1):1–23, 2007. [17] Sonia Ben Mokhtar et al. EASY: Efficient semAntic Service discoverY in pervasive computing environments with QoS and context support . Syst. and Soft., 2008. [18] Naveen Srinivasan, Massimo Paolucci, and Katia Sycara. Adding OWL-S to UDDI, Implementation and Throughput. In SWSWPC 2004, July 2004.