Third International Workshop on Human Factors in Modeling (HuFaMo’18) Silvia Abrahão Miguel Goulão Patrick Heymans Department of Computer Science NOVA LINCS Namur Digital Institute, PReCISE Universitat Politècnica de València Departamento de Informática University of Namur sabrahao@dsic.upv.es Faculdade de Ciências e Tecnologia patrick.heymans@unamur.be Universidade Nova de Lisboa mgoul@fct.unl.pt Xavier Le Pallec Emmanuel Renaux CRIStAL Laboratory Institut Mines-Telecom University of Lille Lille-Douai xavier.le-pallec@univ-lille1.fr emmanuel.renaux@imt-lille-douai.fr I. I NTRODUCTION modeling. MODELS hosted the first two editions of HuFaMo in 2015 Modeling is an intrinsically human endeavour. While con- and 2016. The number of participants, which were between cerned with foundations and technologies, the model-driven 20 and 30, indicates a significant interest on this growing engineering (MDE) community has been somehow neglecting community. The third edition will thus continue to consolidate the issue of human factors in modeling. However, there is a and strengthen it. growing need from the community concerned with quality fac- tors to understand the best practices and systematic approaches II. T HE THIRD EDITION OF H U FA M O to improve the modeller’s experience and confirm the claims of productivity. A particularity of these aspects is that many The third edition of this workshop series (HuFaMo 2018) related questions can only be answered by empirical studies. took place in Copenhagen, Denmark, in October 15, 2018. The HuFaMo workshop is aimed at creating a space for HuFaMo 2018 was held in conjunction with the ACM/IEEE discussion being a get-together of researchers and practitioners 21st International Conference on Model Driven Engineering from different communities including MDE, Usability/UX, Languages and Systems (MODELS 2018), which is the pre- Human Computer Interaction and Empirical Software Engi- mier conference on systems and software modeling. In this neering. third edition, HuFaMo attracted a considerable number of participants, including researchers and practitioners. The work- We perceive MODELS to be a high-quality venue that has shop included the discussion of 6 papers and a working session however not sufficiently reflected on human factors in model- on the setup of an empirical evaluation and its replication at ing in the past. This workshop is an attempt to compensate for different places thanks to the HuFaMo community. what we deem is a major aspect of modeling, as other venues (such as ICSE) have already acknowledged. III. PAPER PRESENTATIONS HuFaMo expressly focuses on human factors, in order to raise the awareness for these topics and the associated The HuFaMo Program Committee selected 6 papers for research questions and methods in the modeling community, presentation in the workshop, representing a spectrum of views providing an outlet for research of this type, guaranteeing high on human factors in software modeling. Here below we briefly quality reviews by people that apply these research methods outline some of the main contributions of each of those papers themselves. Along with fully complete empirical evaluations, and our reflections on them. the workshop organizers explicitly encouraged researchers to Selviandro et al. presented a systematic method to define the discuss study designs before conducting their empirical evalu- concrete syntax of modeling concepts based on the inheritance ations. The rationale was to create a constructive environment structure of their meta-model. The underlying principle is where the HuFaMo participants could contribute to improving named Visual Inheritance. It consists in keeping the visual the proposed study designs so that stronger (and more easily representation of a property - the exact representation or its replicable) empirical designs and results can be obtained. Ulti- design principle - for each subclass (if any) of the class where mately, we aim to congregate a community of researchers and the property is defined. This approach provides a valuable practitioners that promotes (possibly independently replicated) support for creating notations to existing metamodels since empirical assessments on claims related to human factors in it could help the user of the notation inferring the semantic meaning of the classes and reduce the cognitive overload in • Bran Selic, Malina Software Corp., Canada memorising the number of the notations” [3]. • Jean-Claude Tarby, CRIStAL - Centre de Recherche en Silva et al. [6] presented the experimental protocol of an Informatique, Signal et Automatique de Lille, France empirical study that compares two multi-agent systems domain • Juha-Pekka Tolvanen, MetaCase specific languages (DSLs). Generally, people use the Cognitive • Jean Vanderdonckt, Université catholique de Louvain, Dimensions of the Physics of Notations framework to assess Belgium the usability of DSLs. Differently, the underlying motivation of ACKNOWLEDGMENT this experiment is to show that the abstract syntax of modeling languages should be evaluated as well. The organizers would like to thank the authors who submit- Klünder et al. [1] presented an initial attempt to add ted their works to this third edition of the HuFaMo workshop, quantitative analysis capability to FLOW diagrams. This aims all the attendees of the workshop sessions, the PC members to substantiate the results of primarily subjective analyses who reviewed the submissions, and the remaining organization provided by FLOW, with the more objective results drawn members. from qualitative analyses. A tool was implemented extending R EFERENCES the FLOW method, used to analyse and improve the commu- [1] J. Klünder, O. Karras, N. Prenner and K. Schneider, Modeling and nication in software projects. One interview was done in the Analyzing Information Flow in Development Teams as a Pipe System, In industry in a particular case. Third International Workshop on Human Factors in Modeling (HuFaMo Liaskos and Tambosi [2] presented a study design to empir- 2018). CEUR-WS, pages 3-10, 2018. [2] S. Liaskos and W. Tambosi, Comparing the comprehensibility of nu- ically compare qualitative and quantitative contribution links meric versus symbolic contribution labels in goal models: an experi- in goal models according to their intuitiveness and efficiency. mental design, In Third International Workshop on Human Factors in The underlying research questions are how to optimize the Modeling (HuFaMo 2018). CEUR-WS, pages 11-18, 2018. [3] N. Selviandro, T. Kelly and R. Hawkins, Visual Inheritance for De- understanding, the learnability and/or the perception of a signing Visual Notation Based on a Metamodel, In Third International contribution link. Coming from the requirements engineering Workshop on Human Factors in Modeling (HuFaMo 2018). CEUR-WS, community, this work strongly emphasizes human factors and pages 19-26, 2018. [4] E. Renaux, T. De-Wyse and J. Mennesson, Using sketch recognition for shows that a bigger synergy can be established with the capturing developer’s mental models, In Third International Workshop HuFaMo community. on Human Factors in Modeling (HuFaMo 2018). CEUR-WS, pages 27- Renaux et al. presented a software prototype for automat- 34, 2018. [5] J. Lopes, J. Cambeiro and V. Amaral, ModelByVoice - towards a general ically capturing UML diagrams from hand-made sketches. A purpose model editor for blind people, In Third International Workshop lot of knowledge is stored in such sketches because they are on Human Factors in Modeling (HuFaMo 2018). CEUR-WS, pages produced in the middle of a brainstorming session, or while 35–42, 2018. [6] J. Silva, A. Barisic, V. Amaral, M. Goulão, B. Tekin Tezel, O. F. Alaca, explaining parts of the system to some other stakeholder. M. Challenger and G. Kardas, Comparing the Developer Experience Rather than wasting it (generally lost, or thrown away), the with two Multi-Agents Systems DSLs: SEA ML++ and DSML4MAS goal of this work is to benefit from it. Ultimately, the goal is - Study Design, In Third International Workshop on Human Factors in Modeling (HuFaMo 2018). CEUR-WS, pages 43–50, 2018. to “capture efficiently the mental model of the author without asking her/him to transcribe her/his sketch”. [4]. Finally, Lopes et al. [5] presented a software prototype called ModelByVoice that uses voice synthesis and recognition to support visually impaired users in performing modeling tasks. The authors have conducted a preliminary evaluation with blind users to evaluate the effectiveness of the tool. The ultimate goal of ModelByVoice is to improve the accessibility of users by allowing blind people to deal with model-driven development and domain specific modelling languages the same way it is already done with diagrammatic languages in the existing modelling workbenches. IV. P ROGRAM C OMMITTEE • Vasco Amaral, Universidade Nova de Lisboa, Portugal • Arnaud Blouin Insa, INSA Rennes, Inria/IRISA, Diverse Team, France • Michel Chaudron, Chalmers Gothenborg University, Sweden • Cédric Dumoulin, CRIStAL - Centre de Recherche en Informatique, Signal et Automatique de Lille, France • Emilio Insfran, Universitat Politècnica de València, Spain • David Socha, University of Washington, USA