The TrendMD recommendation widget is two lines of code that you install on your website(s). Once installed the widget recommends relevant articles to your readers using our recommendation algorithm.
At least half of the recommendations in the widget at any given time will be from your publications, the remainder (at most half) will be from other publishers in our network.
The TrendMD widget promotes your content:
- on your own article pages (keeping visitors on your sites) - this is free!
- on third party websites (bringing new visitors to your sites) - this costs 1 credit per visitor
The TrendMD widget promotes third party content on your article pages
- every visitor that clicks a link to a 3rd party site earns you 0.5 credits
Learn more about our
credit system and how to
edit content sources for the widget.
In order for the widget to make recommendations the following elements are needed:
Abstract and/or blog content
Title
First author
Publication date
All articles on your site that contain this information will be included in the internal recommendation pool.
The following meta tags are compatible with the TrendMD widget:
Dublin Core
OG meta tags
Highwire Press tags
Eprints tags
BE Press tags
PRISM tags
The following are examples of meta tags we support:
Meta tags for abstract and/or content
Dublin Core meta tag
Meta tag for Title
Dublin Core meta tag
The sponsored status of an article determines where this article will be seen in the TrendMD network.
Sponsored articles will be recommended across the network as long as your credit balance is above zero.
Non-sponsored articles will be recommended in the widget on your website(s) but will not appear in recommendations on other publishers' sites.
You can change the status of your articles on the Professional or Enterprise plans by following these steps.
Similarly, articles that other publishers have designated as 'sponsored' will appear in the widget on your site as third-party links. Find out more about how the widget works here.
In order for TrendMD to generate the initial recommendations for articles, we need to gather key article meta data (information about the article). This allows our algorithm to generate the initial sets recommendations.
Key article meta data includes:
- *Article title
- *Article URL
- *Abstract
- Author names (when available)
- Publication date (when available)
- Journal
TrendMD relies on
crawling to acquire article meta data when a publisher is first onboarded. Once the widget is installed, it automatically picks up meta data from any newly published articles.
* this information is required for the widget to provide recommendations