Engineers in the automotive domain are confronted with a huge load of information making it di cult for them to identify the information relevant for performing their tasks. Particularly challenging is the alignment of process information, such as o ces les, checklists, and guidelines with business processes. In previous work, we introduced the concept of process-oriented information logistics (POIL) enabling the intelligent delivery of process information along business processes. In this paper, we present iGraph, an application implementing POIL. Speci cally, iGraph demonstrates how engineers can be supported with relevant process information during the review of product requirements.
The amount of information engineers are confronted with, makes it a challenging task to identify and handle the exact information needed to perform their daily work. During a review, for example, engineers not only have to consider o ce les and best practices, but also guidelines and handbooks. This information may be accessed through shared drives, databases, or enterprise portals. However, engineers are not only interested in quickly accessing information, but additionally require comprehensive and aggregated information when conducting a review.
To tackle this challenge, information logistics (IL) concepts have been
introduced by researchers and practitioners in recent decades [
German Federal Ministry of Education and Research under grant number 17102X10.
This weakness has guided our development of process-oriented information
logistics (POIL) as a new paradigm for delivering the right process information,
in the right format and quality, at the right place, at the right point in time,
and to the right people [
The core component of POIL is a semantic information network (SIN) [
Section 2 introduces the application scenario. Section 3 presents iGraph. Section 4 discusses related work. Section 5 concludes with a summary. 2
The iGraph scenario deals with the review of product requirements documented as functional speci cations at a large automotive manufacturer. Goal is to both improve and approve such speci cations. The underlying review process is knowledge-intensive as it comprises large amounts of process information, user interaction (e.g., \perform review meeting"), and decision-making (e.g., should the document be approved or not?). Three roles are involved: The author provides the speci cation to be reviewed. The review moderator organizes the review meetings. The reviewer nally analyzes the provided speci cation and documents errors, ambiguities, and uncertainties.
Speci cally, we consider a scenario with one process schema (modeled with Signavio Process Editor), three process instances (created and managed with the Activiti Business Process Management (BPM) Platform), and about 300 documents (i.e., process information) such as reviews, templates, and guidelines. 3
iGraph1 is a web-based Java application based on the semantic middleware iQser GIN server 2.0, the web framework Play! 2.1.1, jQuery 1.8.3, D3 3.1.1, HTML5, and CSS3. The three main features of iGraph are as follows: (1) comprehensive integration of process information and business processes from heterogeneous data sources, (2) intelligent syntactic and semantic analysis of integrated information and process objects, and (3) process-oriented delivery of needed process information and business processes to knowledge workers.
iGraph implements the architectural layers of our POIL framework [
(b) Graph-based view of iGraph.
In the following we show how iGraph is used. Particularly, we consider a speci c task of the review process: the author prepares a functional speci cation for the review and needs a review template for guidance.
Goal is to identify relevant information objects supporting the review preparation. For this purpose, iGraph provides a search box. The reviewer enters, for example, the term \template" into the search box and executes the query. Search results are listed in a table-based view (cf. Fig. 1a). Each row represents a search result (i.e., an information object), whereas each column contains detailed metadata of the found information objects, such as the author or title of an object. In order to identify related information objects (e.g., addressing the same topic), iGraph provides a graph-based view (cf. Fig. 1b) showing related information objects starting from a speci c information object (e.g., a template).
In order to quickly identify relevant information objects, iGraph provides two
fundamental key indicators: the rst algorithm determines the link popularity
(SIN LP for short) of information objects based on the SIN. The second
algorithm determines the rate popularity (SIN RP for short) of information objects
based on user ratings. In [
Michelberger et al.
Various approaches have been proposed in the eld of IL. As examples consider
data warehousing (DWH), business intelligence (BI) solutions, decision support
systems (DSS), and enterprise content management (ECM). However, these
approaches su er from several weaknesses. For example, DWH rather focuses on
the creation of an integrated database. Traditional BI, in turn, addresses data
analytics and is usually isolated from business process execution. Conventional
DSS support complex business decision-making at the management level. By
contrast, ECM deals with the management of information across enterprises
referring to related strategies, methods, and tools. Applications enabling IL are
available, for example, in the elds of wearable computing [
This paper presents iGraph, an application applying semantic technology to enable the integration, analysis, and delivery of process information to knowledge workers. The simple visualization of iGraph, both in a table-based and graphbased view, as well as the two key indicators SIN LP and SIN RP make it easy to identify relevant process information during business process execution.