Resources

Mantis Recommender

Summary, on-boarding steps, documentation and FAQ for Mantis Recommender.

 1.  Mantis Recommender Introduction

Summary

Mantis Showcase is a revolutionary in-page widget powered by the Mantis TAG verified AI Technology that provides seamless, harmonious access to related content from an existing article page, directly driving time-on-site and increasing yield.

Challenge it solves

Page views and time-on-site are the lifeblood of a publisher’s business and the heart of their value-add with consumers, but securing and retaining consistent levels of page views is challenging.

A full one third of users (34%) spent between 1-5 minutes on a news website per month. However, there is a small core of power users that spend more than an hour on-site per month.

This means this small core of power users contributes more than 4x the amount of time on site than casual users (source: Pewresearch, Kenny Olmstead, Amy Mitchell and Tom Rosenstiel 2011 ).

Moving users from one category to the other is vital to maintaining and enhancing yields.

How it works

Mantis’s AI scans hundreds of thousands of articles and classifies them in-context, classifying them in categories, based on the concepts in the article and the entities within them. Showcase then leverages Elastic search to recommend the most relevant content and surfaces the suggestions in a customisable, responsive and intuitive widget.

What it gives you

Showcase has been shown to increase traffic by up to 5%, providing a material impact on revenue, as well as further opportunities for in-widget ad-units. All immune from the tides of cookie-deprecation and without disrupting the flow of your website.

 2. Mantis Recommender Product Workflow

Neptune Recommender will ingest your article text and return in-widget recommendations:

 3. Uploading articles to scan

When an article is to be analysed there are two approaches to integration with Mantis.

• Mantis synchronous HTTP API endpoint
• Mantis Asynchronous Batch File upload and retrieval

These are explained in more detail below.

Mantis Synchronous HTTP API

Call the Mantis ‘classifyArticle’ api endpoint for the customer Mantis account (to be provided) e.g.

https://<publisher-mantis-url>/classifyArticle

 – POST article data in the call to this URL, as a JSON object in the request body including:

• HTML content (just send the article body content itself, not a whole page with headers/footers/links and teasers for other articles/commercial and other embedded components) – or plain text of the article content.
• Public Page URL
• CMS id (publisher specific unique article reference id)
• Author(s)
• Title
• Published/last modified timestamps

 – Pass publisher authentication credentials (to be provided) using an HTTP Basic authentication header

The article will be processed and Mantis Showcase recommendation scores will be generated – and returned from the Mantis API in the response body as JSON data.

In order to analyse multiple articles at once, the endpoint will accept either a single JSON object describing one file, or an array of objects (one for each file to be processed). If an array of article data objects is passed then the response will also be an array of data objects with ratings for each file.  However, it is important to note that this HTTP API operates synchronously and will have to process all articles before a response is sent . It is therefore only suitable for very small batches of files. If larger batches of files are required to be processed then the asynchronous batch file processing interface should be used (see below).

Mantis Asynchronous Batch File Processing

The data formats used to process larger batches of articles are very similar to those used with the synchronous HTTP API endpoints – containing arrays of JSON data objects describing the files to be processed and the ratings for each processed article.

However, in order to accommodate larger batches, the article data is uploaded as separate files to the Mantis platform.

The primary integration method for file upload uses the IBM Cloud Object Storage. A customer-specific storage bucket will be provisioned and the files containing the article data should be uploaded to this bucket.

Details of how to integrate with the IBM Cloud Object Storage can be found here:

https://cloud.ibm.com/docs/services/cloud-object-storage/hmac?topic=cloud-object-storage-upload 

Once the files are uploaded they will be processed by the Mantis analyser and the results will be stored as new files in the same storage bucket – from which they can be downloaded as required.

 4. Receiving your widget script

Within 4 days of successfully scanning articles and confirming what stylistic amends are requested we will share the widget script to be implemented.

We will provide web/iFrame versions as required.

Configuring your widget

The config section of this can be adjusted with the following criteria:

Standard Config

apiUrl: { Customer Specific Recommender Endpoint}

age: {Default Value is 14}

limit: {Default Value is 20},

title: { Default – “SIMILAR ARTICLES TO THIS”},

autoplay: {Default is “false”, to enable or disable auto play of the Slider}

branding: {Customer specific Branding},

Showcase supports Viafoura Comments that can be enabled if required

To enable Tracking

useIntTracking: { Default value is “false”, tracking will be ingested into our shared GA account, this can be used for tracking with internally, can be customisable based on the requirements}

platformType: “web”,

To Enable Ads

recommenderAdsFormat: “all”,

recommenderMaxLbNumber: 2,

recommenderAdsEnabled: ‘true’ {to enable ads}

recommenderAdsFormat: “all”, {Optional: “all” – all types “mpu” – for mpu only; “lb” – for leaderboards only}

autorefreshEnabled: “true” {Optional to enable ads autorefresh}

autorefreshInterval: 40 {Optional to set autorefresh longer then 30 seconds}

recommenderMaxMpuNumber: 10 {to set number of MPU, default: 4}

To fetch Tag specific articles (can be used to run specific campaigns)

queryTag: “”,

campaignTag: “”,

 5. Mantis Recommender FAQs

General

What is Mantis Recommender?
Mantis Recommender is a TAG-verified, AI-powered, on-page widget, showcasing relevant and recent articles to maximise traffic and impressions in a completely cookie-free and customisable manner.

Are there any minimum traffic requirements?
30m page views per month

Can I input ads to the widget?
Yes – contact your Mantis representative to organise integration

How does Mantis Recommender choose what to recommend?
Mantis Recommender leverages IBM Watson’s Natural Language Understanding to comprehensively scan thousands of articles and millions of characters tagging them with multiple variables (categories, concepts, entities, sentiment etc) and scores. Linking these variables together allow us to surface the most relevant and/or recent recommendations possible, all before a page loads.

How effective is Mantis Recommender at driving traffic within site?
We’ve seen significant uplifts of around 2-5%, focussed entirely on driving self-referral within your eco-system.

How does Mantis Recommender work with Mantis Brand Safety?
The recommender is based only on the contextual part of Mantis to maximise the most relevant content. However, if you are signed up to Mantis Brand Safety all of your sites would already be scanned for brand safety and the recommender will load only the most relevant suggestions to that article.

Can Mantis Scan multiple languages?
Yes – in its current version Mantis can scan in English, French, German, Italian, Arabic, Korean, Simplified Chinese, Dutch, Japanese and Brazilian Portuguese.

Will a Mantis Recommender widget slow down site speed?
No, while final performance will vary based on your website, the widget is extremely performant.

Getting started

How do I get started with Mantis Recommender for Publishers?
Reach out to your Mantis representative to arrange an introductory meeting then moving on to discuss your formatting requirements and website architecture.

How do I upload articles to be classified?
Articles can be classified using either a push to our API or via batch upload, details on the mantis-intelligence docs site.

Do Mantis Recommender widgets work on AMP Pages?
Yes

Can I control what shows up on my widget?
In order for Mantis Recommender to maximise the use of it’s powerful recommendation engine we suggest letting it run as freely as possible. We do, however, have controls in place allowing us to exclude certain articles or titles and filter by recency.

What customizations can be made after registering?
Mantis Recommender has been built to allow for complete customization. We can either surface our complete standard widget, adjust for colour and branding or surface the raw API to allow for you to build whatever front-end you require.

Will I have a Mantis Recommender account contact?
Yes

Are there any widget best practices?
We find that Mantis Recommender works best when positioned above the 3rd from last paragraph, ensuring that a consumer has read through the source page before navigating elsewhere.

Performance tracking

How do we track Neptune Recommender CTRs?
Using a joint Google Analytics account we will track CTRs and page views.

Where can I see my performance & reporting?
a joint Google Analytics will be set up upon live-date

Editorials And Control

Is Mantis Recommender SSL compliant?
Yes

What if I see content that I don’t like serving in the widget?
Contact your Mantis representative and we can block the URL or cms id that you would like to occlude.

Can I get a break-down of what categories inform my recommendation eg. This is the top recommendation because of ‘x’ category, ‘y’ entity and ‘z’ concept?
Mantis Recommender uses elastic search to isolate the most relevant articles to reference, this uses a combination of all available categories, entities and concepts and individual element contributions cannot be distilled from the result.