Bibblio is available as monthly subscription packages that are sized to suit your catalogue and usage needs. They start at $129 per month, with a 60-day refund option. You can also try Bibblio for free on a pilot plan.
A content item is any article or page within your catalog that you want to provide recommendations to and from. Sets of metadata are collected for each content item, which is then stored in our Smart Graph. Depending on your plan, there’s a limit to the number of content items that can be stored at any time.
A recommendation call is a request to the discovery endpoint of the Bibblio API and subsequent delivery of recommendation suggestions for the content item you’re currently viewing.
No, there are no hidden costs to sign up.
We accept all major debit and credit cards.
If you have forgotten your password you will need to reset it. You can do so here.
Not currently, but we’re looking into it.
Not currently, but watch this space.
We can produce recommendations for all languages using the Latin alphabet, and we can generate in-depth semantic analysis for English and Spanish.
Currently Bibblio can only enrich text, but we’re working on accommodating speech and video in the near future. Get in touch if you would like to be involved in a preliminary speech or video project with us.
At the moment, content needs to be formatted in valid JSON to be pushed to our API, but we’re currently working on tools to help with URL scraping and pulling directly from RSS feeds, as well as developing a plugin for all WordPress publishers.
The maximum size for an individual item of content is currently 200 KB.
Our documentation generates basic scaffold code for each endpoint in a wide variety of languages. You can find it by switching to the console mode on an endpoint, calling the resource, clicking Code Example, then selecting your language.
There is an Xblock available to allow you to integrate with the edX course platform with minimum fuss. This was built by our good friends at Proversity.
Finally, a Wordpress plugin is in the works. Stay tuned.
A 422 is often because the payload is invalid. For example, required fields could be omitted or dates might be incorrectly formatted.
If that doesn’t seem to be the case, please get in touch.
Metadata is data about data. Bibblio’s metadata is split into structural metadata about the data container (the file), and descriptive metadata about the content itself, such as the relevant keywords in an article. Descriptive metadata is what Bibblio creates, and it serves as the basis of our recommendations.
If you want to dig a little deeper, check out the Wikipedia entry on metadata.
Metadata is automatically added to the content items, but it won’t appear as tags on your web page or platform unless you engineer that.
In order to make the tags generally applicable, we don’t match tags to individual client taxonomies. Any filtering of Bibblio’s tags based on a pre-existing taxonomy would have to be accomplished through logic on your side.
You may not have supplied enough text to derive concepts, keywords, entities and alignments. (Generally, a few lines of text should be sufficient.) Alternatively, the text might be in a language that our enrichment endpoint does not currently support.
Your access token is probably not valid anymore. We mitigate the risk of exposing user credentials by using a temporary token according to the OAuth2 Bearer Token authentication mechanism. You'll notice an expires_in field in the token response. Clients generally create an authorization wrapper that renews the access token according to the expires_in value provided by the token endpoint. Learn how to get a new token.
If you’re experiencing this problem while using a valid access token, please get in touch.
When we speak about discovery, we mean the process through which a user finds and interacts with content, whether that’s by traditional search, on-platform content recommendations, or in the form of a personalized newsletter.
Bibblio does not drop any cookies on the page. All user tracking is carried out from within the related content module itself, or is sent to us directly by our partners.
No, and they never will.
Yes, our recommendations are tailored to the user they’re presented to. We integrate user session data with our in-depth content analysis approach to create a more personalized recommendation experience.
bower install bibblio-related-content-module
They do! Our modules are fully responsive, collapsing gracefully for display on the smallest devices.
You can use the recommendations however you like, but our elegantly designed modules are ready to go and fully customisable. For most clients they make getting recommendations a lot more convenient. In addition, we also track user behaviour directly through our module.
We offer support across all major browsers (Chrome, Firefox, Safari etc.).
The module does not block rendering. The module populates an HTML element of your choice with content. An empty <div> (or any tag) can be rendered quickly, and the module can be initiated to fill it with content at any point during page load, such as page ready. Any delay in loading the module content will not affect the rest of the page in any way.
We store the source URL and module image URL when you add the content, so we can easily populate the module and don’t double up on server storage. When the user clicks, we simply send the user to the original content URL.
Content items can be grouped into what we call ‘catalogues,’ which means you can organise your content to ensure that certain topics and genres are only recommended on certain parts of your platform.
We are in the process of enabling 'open catalogues' which means that anyone using Bibblio’s service will be able to access content from our curated open catalogues or from other Bibblio clients who open their catalogues. As a client, you will also be able to distribute your content to other users.
Our dashboard allows you to see how many enriched content items you have stored in Bibblio’s Smart Graph and how many recommendations have been served, both monthly and in total.
The latency of recommendation delivery via our modules is below 400 milliseconds, which is often quicker than the other page elements on our clients’ platforms.
On average, you can expect recommendations to take between 150ms and 1000ms, and the response times for content item ingestions are roughly the same (from the UK).