Artificial intelligence is changing the publishing industry. It’s speeding up publication and improving data-processing capabilities. Moreover, AI tools can benefit researchers too. Here, we’ll explore the AI tools that MDPI’s AI team are developing for researchers. Whilst doing so, we’ll also introduce you to the AI Team.

MDPI’s AI Team

In the world of scientific research and publishing, AI is being implemented to improve processes, services, and the quality of research. MDPI is committed to being at the forefront of Open Access scientific publishing, so it is very interested in exploring how it can leverage AI to improve MDPI’s services.

Over the last year, the AI team has worked on setting up the foundational pipeline that is now allowing us to develop and launch different tools. This has involved collaborating with departments across MDPI and studying the broader scientific community to identify key areas to focus on.

On the one hand, the AI Team is aiming to improve the efficiency of the workforce by increasing the efficiency of the workforce and simplifying time-consuming tasks. On the other hand, they are also deploying tools to improve the user experience of authors, reviewers, and readers.

Throughout this article, we’ll introduce you to the team and the work they’re doing.

MDPI’s AI tools for researchers

AI is dramatically shifting scientific publishing. A previous article explored the range of advantages and disadvantages that have emerged from this change.

The advantages include data management and task automation, translation aids, and image generation. The disadvantages, however, include difficulty in detecting generative AI and fake data, the race between tools to detect AI and growing AI sophistication, and paper mills. To learn more about these issues, read the full article Artificial Intelligence in Scientific Publishing.

How MDPI’s AI Team create tools

Diogo Rodrigues, Senior Data Scientist, explains how the team produces AI tools for researchers and more:

As part of the AI team, we’re committed to creating user-centric tools for both internal and external users. Our primary objective is to deliver valuable applications that optimize workflows such as publishing research and providing support through features like identifying related works or recommending appropriate journals.

To ensure state-of-the-art performance, our team incorporates cutting-edge technologies and innovations, including large language models. We recognize the elevated standards within the scholarly community and strive to exceed your expectations.

Collaboration is key to our success. So, by working closely with business teams, we maintain alignment with current developments while ensuring future offerings add significant value to users. Seamless communication between domain experts guarantees that our internal stakeholders and, ultimately, researchers profit from our applications.

AI tools for researchers

Here is an overview of some of the AI tools that MDPI’s AI Team have developed.

Journal Finder

The Journal Finder tool provides a selection of the most appropriate journals for you to submit to based on the title and abstract of the paper. The tool is also used internally to verify that the submitted manuscript fits the scope of the journal.

This tool takes the title and abstract and passes them through a natural language processing (NLP) model that generates a numerical representation of the text that can capture its meaning (this is called an embedding). The embedding is then used to do a similarity search among the papers in each journal to find papers that are similar in scope to the submitted manuscript. It looks at patterns within the language of your work and the content in journals, rather than just keywords.

Furthermore, the tool provides a filter to narrow down the search, based on journals’ Impact Factors and the database index.

Article Similarity

Article Similarity is an AI engine that suggests similar or related papers based on the one you are viewing. The tool is integrated with mdpi.com, Scilit, and Preprints.

It can be used to find articles with a similar topic regardless of the journal, ultimately increasing the discoverability of relevant research. Further, it can be used for peer reviewer recommendations, topic modelling projects, and more.

Category-connect

MDPI developed Category-connect, which leverages AI to categorise scientific papers into different research areas. By using an AI algorithm, the tool provides relevant categories for manuscripts. This is like author keywords, but in a more consistent and standardised manner.

This powerful tool allows the editorial staff to have an overview of published papers and help them to make data-driven decisions.

For authors, the tool will provide higher visibility to MDPI papers and allow researchers to stay up to date with active research in their field. For readers, it will help in finding related papers published by MDPI, aggregated in Scilit, and archived in Preprints.

Ethicality

Ethicality is an AI tool designed to help with the ethical checks during the manuscript submission process for MDPI’s journals.

Thanks to the experience and professionalism of the editors and ethical team, the AI Team were able to gather different examples of potential ethical issues that will be used in the training dataset.

It will be fully integrated into MDPI’s online submission system, SuSy, and will detect and flag potential ethical issues during the technical pre-check phase for new submissions. This will allow Assistant Editors to quickly check the potential issues and resolve them. Also, it will add valuable information on the submission level that can be checked at any time whilst processing the manuscript.

Furthermore, it features a reference checking function. This feature will automatically search for and, if present, highlight several potential issues with the references, such as self-citation rates, non-related citations, old references, etc.

AI tools and MDPI

The world of artificial intelligence is very vast and dynamic. MDPI is committed to being at the forefront of Open Access scientific publishing. MDPI believes that AI can provide innovative solutions and enhance the experience of interacting with MDPI material. The AI Team are producing AI tools for researchers that reflect the values of the company and, more importantly, benefit the whole scientific community.

Diogo Rodrigues, again, summarises MDPI’s ambition with AI:

Our main ambition? To create intuitive tools that simplify author tasks, empowering them to concentrate on what truly matters: advancing research and fostering a better, more sustainable world. With modern technology and active collaboration with domain experts, we continually commit ourselves to develop solutions that support academics.

If you are interested in submitting research and integrating some of these tools into your workflow, have a look at the full journal list.

This article was written in collaboration with Enric Sayas, Business Analyst for the AI Team.