Searching textual and model-based process descriptions based on a unified data format

Abstract

Documenting business processes using process models is common practice in many organizations. However, not all process information is best captured in process models. Hence, many organizations complement these models with textual descriptions that specify additional details. The problem with this supplementary use of textual descriptions is that existing techniques for automatically searching process repositories are limited to process models. They are not capable of taking the information from textual descriptions into account and, therefore, provide incomplete search results. In this paper, we address this problem and propose a technique that is capable of searching textual as well as model-based process descriptions. It automatically extracts activity-related and behavioral information from both descriptions types and stores it in a unified data format. An evaluation with a large Austrian bank demonstrates that the additional consideration of textual descriptions allows us to identify more relevant processes from a repository.

Missing Procedural Texts Repairing Based on Process Model and Activity Description Templates

“…A number of representative work has been done. For example, [24] proposed to extract relevant activity and behavior information from textual and model descriptions and store them in a uniform data format to achieve alignment; [25] proposed to extract the language features of two description types and use integer linear programming (ILP) technique to complete the alignment between the two representations; [26] proposed to convert textual and process model activities into quantifiable word bags, and used the proposed algorithm to calculate similarity to achieve alignment and determine inconsistency. Following [26], [27] proposed a more comprehensive alignment method, including process model activities, events and gateways.…”

Extracting Declarative Process Models from Natural Language

“…Textual documents, such as work instructions and process descriptions, represent a valuable source of information in the context of business process management [3]. Their value has led to the recent development of a variety of analysis techniques that focus on process information contained in text, such as techniques that compare process models against textual descriptions [1,29], as well as techniques that consider textual descriptions for process querying [22,23], process matching [6,33], and conformance checking [2]. Closely related to the goal of this paper are existing techniques that focus on the extraction of imperative process models from natural language texts [15,16,27], of which the technique by Friedrich et al [15] can be regarded as the state-of-the-art in this context [27].…”

Aligning textual and model-based process descriptions

“…However, these approaches have been found to produce inaccurate models, which require extensive manual revision [39]. Other use cases involving texts include a technique that considers work instructions when querying process repositories [25] for conformance checking against textual process descriptions [49].…”

“…These techniques have been widely applied in the context of textual process description for the extraction of activities and their actors, cf. [13,47,25]. We can employ such existing techniques in order to extract actions, actors, and business objects from a text.…”

“…Furthermore, the authors of the text were not involved in the development of the approach of this paper nor aware of the exact purpose of their task. For the 25 model-text pairs in this new collection, the gold standard alignments were annotated and subsequently verified by the authors of this paper. Table 2 provides an overview of the characteristics of the model-text pairs included in the evaluation set.…”