At MDPI, we strive to make your online publication process seamless and efficient. To achieve this, our team is continuously developing tools and features to make the user experience useful and convenient.

As the number of academic papers continues to grow, so does the need to analyze and work with them on a large scale. This prompted us to design a new feature aimed at helping researchers find journals that are relevant to their publication by matching their abstract topic. In this regard, we designed a similarity model that automatically identifies the most suitable academic journals for your paper.

We are pleased to introduce Journal Selector, a new feature that measures similarity in academic contexts. By simply entering the title and/or abstract into our Journal Selector, the author will see a list of the most related scientific journals published by MDPI. This method helps authors select the correct journals for their papers, highlighting the time of publication and citability.

The methodology is known as representation learning, where words are represented as vectors in hyperspace. Representation helps us differentiate between different concepts within articles, and in turn, helps us identify similarities between them.

We used an advanced machine learning model to better capture the semantic meanings of words. This helps the algorithm make better predictions by leveraging scientific text representation. In turn, this ensures high precision, helping authors decide which journal they should submit their paper to.

The goal is to support authors to publish their work in the most suitable journal for their research, as fast as possible, accelerating their career progress.

Contact: Andrea Perlato, Head of Data Analytics, MDPI (email)