Scholarly publishing has been on the rise for many years, but the scale of scholarly output has gone unnoticed in recent times. According to an estimate by Science.org, the number of research papers being published grew to a total of 2.82 million in 2022, from 1.92 million in 2016. So, how can publishers cope with the quantum of publishing activities that are growing exponentially every year? In this blog post, we aim to highlight the intricacies of managing submissions through technology.

 Every author submission is meticulously assessed and evaluated by a publisher before being passed on for peer review. The average time from submission to publication was placed at 163 days in 2020, according to an estimate by the Scholarly Kitchen. The duration suggests significant room to improve the efficiency of research paper reviews.

 Hundreds of scholarly publishers currently operate across the world ranging from a single journal to several hundred, but they all face the same challenges. So, how can an author be confident about the readiness of their papers for publication? Let’s take a closer look at the activities that constitute readiness for publication:

• Correct format: To ensure research papers are publication-ready, authors must adhere to the publisher’s specific formatting guidelines, including font type, size, spacing, and page margins. Consistent use of headings, subheadings, and numbering is also essential. Accurate citation of sources using the required style (e.g., APA, MLA, Chicago) is another crucial aspect.

• Correct references: Complete and accurate bibliographic information for all cited sources, including author names, publication titles, dates, and publication details, is necessary. Proper in-text citations are also required to indicate where information from sources is used. A well-organised reference list or bibliography should include all cited sources at the end of the paper.

• The Correct structure: A clear and logical organisation of ideas is essential for a publication-ready paper. A well-defined introduction, body, and conclusion provide a robust framework. A coherent flow of thought, with smooth transitions between paragraphs and sections, ensures the reader can easily follow the argument. Effective use of headings and subheadings guides the reader through the paper and helps to clarify the main points.

 Traditionally, all of the above was checked by hand, which wasn’t a problem when the number of journals was relatively low. Today the sheer quantity of submissions makes it a challenge to process them manually. Clearly, it makes sense to use automation wherever possible to assist with such a vast undertaking.

 Fortunately, there are many tools available to assist publishers assess and evaluate a manuscript. A surprising number of tasks can be done by a machine; basically, machines are good at doing repetitive tasks such as matching and counting. Checking that all the figure labels are in the correct order is one such check. Checking that all the citations at the end of the paper are actually referred to in the text is another.  Here is a list of some of checks that Molecular Connections provides:

• Article structure: is there an abstract? Is there a methods section? Does the article end with a conclusion?

• Conflict of interest: have the authors included a conflict of interest statement? Does the article include content that suggests a conflict of interest statement might be relevant?

• Counts: the system can count the number of words, number of figures and tables, number of references, and ensure that each reference is cited in the paper.

• Compliance with journal requirements: is the title too long (or too short)? Many journals have maximum word lengths, or maximum of characters.

• Are the subject keywords provided by the author the most relevant?

• Author attribution: are all the author attributions provided, and do they each have an ORCID ID?

 Conclusion

It is clear from the above that there are many gains from running technical checks before any human assessment. Most importantly, the system provides checks but does not make any changes to the manuscript. The author or the publisher can then make changes as they see fit. In other words, this system provides the best of both worlds: a machine identifies the problems (and may suggest a solution), but it is the authors or the editors who make the final decision.