About ASReview

ASReview uses state-of-the-art active learning techniques to solve one of the most interesting challenges in screening large amounts of text: there’s not enough time to read everything!

With the emergence of online publishing, the number of scientific papers and policy reports on any topic is skyrocketing. Simultaneously, the public press and social media also produce data by the second. Suppose you are writing a systematic review or meta-analysis, updating a medical guideline, developing evidence-based policy, or scouting for new technologies. In that case, you need to systematically search for potentially relevant documents to provide a comprehensive overview. ASReview, developed at Utrecht University, helps scholars and practitioners to get an overview of the most relevant records for their work as efficiently as possible while being transparent in the process. It allows multiple machine learning models, and ships with exploration and simulation modes, which are especially useful for comparing and designing algorithms. Furthermore, it is intended to be easily extensible, allowing third parties to add modules that enhance the pipeline with new models, data, and other extensions.

The ASReview Team

The ASReview project has snowballed, and it can be hard to distinguish activities from each other and figure out who is involved in what part. Additionally, it can be helpful to know by whom decisions are made concerning certain aspects of ASReview. In this section, we provide an overview of the organizational structure of ASReview and the responsibilities for decisions made regarding specific aspects of the project.

The People

One of the five key principles of ASReview is that people are the oracles, making the final decision. Therefore, it is fitting to first describe the humans in the project and how we define certain (sub)groups of people in ASReview.

Community

The community consists of all people who are in some form interacting with ASReview and with each other. Specific parts of the project can have different communities. For instance, not all users of ASReview LAB will use the Benchmark Platform, and vice versa.

Contributors

Contributors are people who helped to develop ASReview. For instance, a contributor can add code to certain parts of the software, answer questions on the discussion board or provide teaching materials for others to learn how to work with ASReview. A complete list of code contributors can be found on the contributors’ page on Github.

Maintainers

Who is aware of almost everything in this part of the project? Who oversees what code gets implemented in the software? The maintainers are the leaders of that piece of software. They are the ones contributors look to for answers to their questions and who are responsible for merging pull requests. For example, Yongchao Ma and Jonathan de Bruin are the maintainers for ASReview LAB.

Project leadership

ASReview is an open-source project coordinated by Rens van de Schoot (full Professor at the Department of Methodology & Statistics and ambassador of the focus area Applied Data Science at Utrecht University, The Netherlands), together with Jonathan de Bruin (Lead engineer of the ASReview project and working at the Information and Technology Services department at Utrecht University). Rens serves as the project’s academic lead and Jonathan as the project’s technical lead.

Advisory board

The advisory board consists of machine learning expert Daniel Oberski, (full Professor of health and social data science at the Department of Biostatistics at the Julius Center, University Medical Center Utrecht), Lars Tummers (full Professor of Public Management and Behavior at Utrecht University), Ayoub Bagheri (NLP-expert at Utrecht University), Bianca Kramer (Open Science expert at the Utrecht University library), Jan de Boer and Felix Weijdema (Information specialists at the Utrecht University library), and Martijn Huijts and Carolin Mödinger (UX-experts at the department Test and Quality Services at Utrecht University).

Modules

ASReview is separated into different modules that have a hierarchical structure. All modules fall under the ASReview project itself, ensuring that we keep the different aspects of the project as much aligned as possible. There are three key modules, Software, Research and Education. Within those modules, multiple activities take place that are related to ASReview. Some of those activities fall under the ASReview project, others are appreciated and acknowledged, but fall outside of any decision structure.

Software

Under the software module fall activities such as ASReview LAB, the Benchmark Platform, and plug-ins and extensions that are included within the ASReview github repository. For all this code within the official github repository the project’s technical lead is the main responsible person. In pactice, the maintainers of the activities are the first leaders with gratitude for all input and help by all contributors and the full community. We would like to acknowledge that there are also many wonderful plug-ins and extensions that fall outside this scope and want to thank the community in driving these developments!

Research

With respect to the research, we want to make a division between work that supported the development and proof-of-concept of ASReview up-to version 1.0 and work that is related to ASReview but in other ways. Work related to the former is included in the github repository and serves as legacy to show upon which research the product was build. In that period, the academic lead was “in-charge” of this type of research. Now that ASReview has matured, many opportunities for research are open, only limited by our own imagination. You, as community, are in charge of what happens here! We encourage anyone to share findings and work related to ASReview. None of this new research will however be incorporated within the github repository, unless it leads to new code for the software itself.

Education

Getting started with ASReview by finding relevant materials or visiting a workshop. Education is always very important for the use of any software. The education that is offered through ASReview Academy will fall under the supervision of the academic lead of the project with the teaching team in charge of specific activities. We welcome anyone else to also start teaching with and about ASReview. Please share with us your experiences and share what works well. Just like with the research, you as community in charge and we hope to see many initiatives thrive.

What about

Artwork

Regarding the artwork for ASReview we would like to give some special attention. The original artwork for the ASReview project is created by Joukje Willemsen.

Interactions

Please be aware of the Code of Conduct for ASReview. We expect everyone involved in the project to adhere to the standards outlined in the document. We want to not only be open-source software, but also an open and welcoming community for everyone. In addition, here you can find the research principles to which we adhere.

Citing ASReview

The paper published in Nature Machine Intelligence can be used to cite the ASReview project.

For citing the software ASReview Lab, refer to the specific release of the software. The menu on the right (in Zenodo) can be used to find the citation format of prevalence.

For citing the documentation (or to download the pdf) go to Zenodo.