ASReview Class 101
This blogpost is meant to give insight into the most important aspects of ASReview LAB. It…
Although the software is for free (and Open Source), the development is not... please donate and support the development!
Although the software is for free (and Open Source), the development is not... please donate and support the development!
Anyone who goes through the process of screening large amounts of texts such as newspapers, scientific abstracts for a systematic review, or ancient texts, knows how labor intensive this can be. With the rapidly evolving field of Artificial Intelligence (AI), the large amount of manual work can be reduced or even completely replaced by software using active learning.
By using our AI-aided tool, you can not only save time, but you can also increase the quality of your screening process. ASReview enables you to screen more texts than the traditional way of screening in the same amount of time. Which means that you can achieve a higher quality than when you would have used the traditional approach.
Consider the example of systematic reviews, which are “top of the bill” in research. However, the number of scientific papers on any topic is skyrocketing. Since it is of crucial importance for the advancement of science to produce high-quality systematic review articles, sometimes as quickly as possible in times of crisis, we need to find a way to effectively automate this screening process. Before Elas* was there to help you, systematic reviewing was an exhaustive task, often very boring.
*Your Electronic Learning Assistant who comes with ASReview, read more about The principles of Elas.
Traditionally, the pipeline of a classical systematic review starts with the reviewer doing a keyword search to retrieve all potentially relevant references.
The reviewer can then start screening the abstracts and titles to assess the potential relevance to his/her particular research question.
For an experienced reviewer, it takes between 30 seconds to a couple of minutes to classify a single abstract, which easily results in hundreds of hours spent merely on abstract screening.
After all of the abstracts have been screened by the reviewer, the result is a subset of the initial search containing all potentially relevant references. The reviewer reads all of the full-text versions and writes his awesome paper.
Traditionally, the pipeline of a classical systematic review starts with the reviewer doing a keyword search to retrieve all potentially relevant references.
The reviewer can then start screening the abstracts and titles to assess the potential relevance to his/her particular research question.
For an experienced reviewer, it takes between 30 seconds to a couple of minutes to classify a single abstract, which easily results in hundreds of hours spent merely on abstract screening.
After all of the abstracts have been screened by the reviewer, the result is a subset of the initial search containing all potentially relevant references. The reviewer reads all of the full-text versions and writes his awesome paper.
Similarly, in the research cycle of a systematic review with machine-aided systematic reviewing, the reviewer also starts with a keyword search to retrieve all potentially relevant references, downloads these and imports them into a reference manager.
Then, the reviewer selects some relevant target papers. A machine learning model is trained on these papers to predict which reference to present next. Subsequently, the reviewer enters the active learning cycle.
Since the abstracts are presented from most to least relevant, the reviewer can stop reviewing after having seen all relevant abstracts. This form of active learning will save hundreds or hours time going through all references.
Since we embrace Open Science, all decisions made by the reviewer as well as all the technical information is stored in a log-file, which can (or should) be published alongside the paper.
Similarly, in the research cycle of a systematic review with machine-aided systematic reviewing, the reviewer also starts with a keyword search to retrieve all potentially relevant references, downloads these and imports them into a reference manager.
Then, the reviewer selects some relevant target papers. A machine learning model is trained on these papers to predict which reference to present next. Subsequently, the reviewer enters the active learning cycle.
Since the abstracts are presented from most to least relevant, the reviewer can stop reviewing after having seen all relevant abstracts. This form of active learning will save hundreds or hours time going through all references.
Since we embrace Open Science, all decisions made by the reviewer as well as all the technical information is stored in a log-file, which can (or should) be published alongside the paper.
Our team is working hard and we are continuously improving our software, and we are very proud of what we can already offer you.
Although the software is for free (and Open Source), the development is not… please donate and support the development!
A team of dedicated researchers, engineers, and information specialists are working on the ASReview software. Research and development are very time-consuming. For the nice-to-have features of ASReview, we need your help! You can help not only by developing new features yourself or by improving manuals on Github, but also with a (small or large) financial contribution.
The following documentation will guide you in installing the ASReview software.
Step-by-step installation guide to install the ASReview software on your Windows machine
Step-by-step installation guide to install the ASReview software on your MacOS machine
We will keep you updated with our adventures.
This blogpost is meant to give insight into the most important aspects of ASReview LAB. It…
If you are screening in ASReview, you may want to find more information on the title…
How A.I. may make sense of 50,000 coronavirus research papers BY JONATHAN VANIAN April 28, 2020…
The ASReview tool offers effective, reliable and transparent title-abstract screening. It combines machine (deep) learning…
Hi! My name is Joukje Willemsen. I’m the creator of ELAS (the icon of ASReview) and…
The call ‘Corona: fast-track data’ is developed by the NWO to provide funding for projects that…
Jahoe! We started working in our new innovation lab!!! You wouldn’t necessarily expect to see them…
The ASReview team wins the KNVI- Victorine van Schaickfonds Initiative Award!!! We are very happy with this…
ASReview is a research project and here we present our publications.
After manually screening 5050 studies, Smid et al included only 27 studies for their review. In…
To determine the defaults we performed a simulation study and Naive Bayes + TF-IDF model performed…
We show that by using active learning, ASReview can lead to far more efficient reviewing than…
The CORD-19 database is available in ASReview and can be used to search for relevant Corona-related…
The idea of combining human intelligence and machine learning can be found in machine learning techniques…