✨ Communicating Research Findings to Practitioners - Episode 4 (Weekly) - This one is on a new approach that can be used for specifying Machine Learning Systems that made a successful move from academia to industry! ✨
🚀 I am excited to share the research led by my former Ph.D. student Hugo Guarín Villamizar at Departamento de Informática PUC-Rio on enhancing the specification process of machine learning (ML)-enabled systems. 🤖💡
The paper "Identifying Concerns When Specifying Machine Learning-Enabled Systems: A Perspective-Based Approach," just accepted for publication in the Journal of Systems and Software, one of the most prestigious scientific journals within our field, presents PerSpecML. This novel perspective-based approach provides a solution for specifying ML-enabled systems. PerSpecML is freely available and is already being used by companies to bridge the communication gap between business stakeholders and technical teams, ensuring that ML projects align closely with business objectives and user experiences while leveraging optimal infrastructure, models, and data strategies.
The freely available ML specification template ( https://lnkd.in/dyfpcapA ) is being used at ExACTa PUC-Rio, by our industry partners and by several other companies! It quickly passed 1000 unique views on Miroverse and has been copied for usage by more than 200 different users.
Key Benefits:
1️⃣ Improved Clarity and Collaboration: PerSpecML enables teams to identify critical attributes across ML and non-ML components, fostering better understanding and cooperation.
2️⃣ Enhanced Decision-Making: By offering a comprehensive view of system specifications, the approach aids in making informed choices about trade-offs and priorities.
3️⃣ Unveiling Hidden Requirements: The methodology excels at revealing overlooked requirements, thereby reducing the risk of missing out on essential system components.
🔍 The validation of PerSpecML across academia and industry showcases its potential to significantly improve the specification of ML-enabled systems, making it a promising tool for practitioners in the field.
As we continue to integrate ML into various aspects of technology and business, approaches like PerSpecML are crucial for ensuring these systems are not only innovative but also aligned with our broader goals and values!
🔗 For more details on PerSpecML and its impact, check out the authors' version of the full paper in the comments.
Authors: Hugo Guarín Villamizar , Marcos Kalinowski , Hélio Côrtes Vieira Lopes, and Daniel Mendez.
P.S. Figure generated with DALL-E.
#MachineLearning #DataScience #SoftwareEngineering #ResearchInnovation