gasexchange

To welcome in the new year, metajournal quietly rolled-out a big update to the machine learning magic that is the foundation for recommending quality, significant and personally relevant papers to you based on your professional medical interests.

The metajournal Collective Intelligence Engine, is the secret sauce that integrates evidence with your own interests, along with topics that other clinicians in your specialty are following and interested in.

Not only does metajournal look for quality articles that are relevant to the Topics and Journals that you follow – metajournal is now smart enough to follow your article reading behavior.

As you read and favorite papers and abstracts, metajournal learns more about the types of articles that you find most interesting and relevant – the topics, the authors, the journals, and many many more features. Metajournal will even recommend particularly relevant papers that are outside your followed journals, reducing the chance that you miss something truly important.

Metajournal is also smart enough to understand which articles you do not find interesting – simply click the not interesting... link under an abstract and metajournal will grow a little wiser by understanding what is not relevant to you.

The best way to get the most out of metajournal's growing intelligence is simply to teach metajournal more about what interests you: follow interesting Topics, click through to Read interesting abstracts and Favorite the most interesting papers, and when metajournal suggests Not interesting... articles – click the link to tell it!

When browsing your articles page you may have noticed a new link: trending articles.

Trending articles gives you a glimpse into which papers the rest of the metajournal community, with similar clinical interests to you, are reading, sharing and studying.

These are not necessarily the latest or most recent articles, but rather those that currently most interest metajournal users who follow the same journals as you.

Next to the trending tab, you will also find a link to the latest articles - showing the newest articles published from all the journals that you follow in your metajournal account. This is a great way to dip your toe into the waters of the latest published research in your field, browsing through the freshest and newest papers.

Between metajournal's article suggestions, trending and latest articles, you will be certain to find something that not only interests you – but also improves your practice.

Metajournal now suggests related articles underneath the abstract you are reading.

This is a simple feature that may have a dramatic impact on your understanding and appreciation of the greater body of evidence surrounding a topic.

Metajournal goes a step further than PubMed's suggestions of related articles, by not just considering the similarity of titles and abstracts, but by also leveraging the machine learning of the metajournal Collective Intelligence Engine to consider article significance, relevance and quality.

For nice examples of Related Articles in action, have a look at the SAFE study (NEJM 2004) and the B-Aware trial (Lancet 2004) – where the suggested 'related articles' include both interesting papers and more recent studies that may otherwise not be seen.

Metajournal has extensive search features allowing logged-in users to search within article titles – either from among the entire 1,680,360-articles index, or just a smaller slice.

Simply click Search... and enter your search terms in the box. Search within your Suggested Articles, your Favorites, articles from one Topic, or among articles from just one Journal.

From there you can widen the search to metajournal's entire article index, or then even perform the same search on PubMed.

And if that's not enough, you can also search metajournal's Collections and Notes too!

You can easily follow extra journals with metajournal, and so increase the range of publications used to recommend personally-relevant articles.

Sometimes the most important specialty-specific articles are published in the big journals, such as The Lancet or The New England Journal of Medicine – and now metajournal indexes articles from these journals, as well as from JAMA, The BMJ, CMAJ and MJA. (You can also directly browse articles from the journals themselves.)

The easiest way to add extra journals to follow is via your settings page. Click on Journals and from there you can easily search for new journals to follow by name, or unfollow journals that no longer interest you.

When you first join metajournal, you automatically follow all the main publications from your specialty. However by refining your list of followed journals you can fine tune things further.

If you follow journals outside your specialty, its important to tell metajournal what your clinical interests are by following a good number of Topics (at least ten). Metajournal uses the topics that you follow to predict and suggest specific articles that are relevant to your practice, even if they are published in journals outside your specialty.