Xeerpa Recommender:

In order to comply with certain social networks' use cases and policies (namely Facebook, Instagram and Google), websites implemmeting social login with these networks must implement a personalised content recomendations feature. The Xeerpa Recommender widget provides such functionality.

It consists of a content slider or mosaic which is returned by Xeerpa as a JSON string with HTML code, showing relevant content recommendations based on all the information the user has accepted to share with the brand. The HTML can be inserted in an empty  div element and requires jQuery and UTF-8 encoding.

Currently, the widget can recommend content based on: likes and preferences, geolocations, posts, gender and age. These content recommendations must be provided by the client so it can be configured in the widget by the Xeerpa team.

The recommended content consists of:

- One working URL for each recommended content

- One 580x290px / 72ppp resolution relevant image for each URL

- Recommended content should include at least:

• Two slides for gender (one for male users, one for female users)
• Three slides for age (for different over 18 generations)
• Five slides for interests (one for each Xeerpa categories group: "Household", "Leisure", "Art & Beauty", "Industry & Digital World", "Lifestyle")
• As many slides as possible for geolocations. These should recommend places to visit which are relevant to the client and close to user's location (i.e. stores, bars, stadiums, etc.)

This is how the Xeerpa Recommender looks like with some content examples:

Home screen:

Age based content:

Gender based content:

Geolocation based content:

Likes/posts based content:

Technical details:

How to implement the Xeerpa Recommender:

Using the Xeerpa Connect Back-Auth method

https://auth-CLIENTURL/widgets/summary?id=AU_ID&token=AU_TOKEN&lang=LANG&socialNetwork=SOCIALNETWORK

Using the Xeerpa Connect Discovery method

Xeerpa Connect Widgets will return a JSON string with a parameter called orwith the code to embed into your app:

A successful response would be a JSON with:

An example of a call for the Xeerpa Recommender through Xeerpa Connect BackAuth looks like this:

An example of an URL through Xeerpa Connect Discovery including the call for the Xeerpa Recommender looks like this:

*Xeerpa Recommender can also be requested through a request to /discoveruser endpoint, adding connect={"show":"summary"} to the query.

Proximity Widget:

The Proximity Widget will recommend stores, bars, or locations from a list predefined by the Client after the user logs in, based on the user's locations history.

You can customize the parameters of the call with the specific parameters of the Proximity Widget:

- PRECONFIGURATION IN DASHBOARD:

Xeerpa will need a JSON with this structure:

Address and fbid are Optional parameters.

YearsAgo is used to establish the maximum antiquity of the location.

Each item.loc must be in <long, lat> format.

 - OUTPUT:

After the user logs in, a JSON array will be generated with the following format:

Top recommendations will be sorted by class (recommendation level). The class will depend on the density of the locs of the user, near a specific location.

These classes can be A-E.

An example of a call for the Proximity Widget through Xeerpa Connect Backend Auth PHPs looks like this:

An example of an URL through Xeerpa Connect Discovery including the call for the Proximity Widget looks like this:

Engagement Widget: 

The Engagement Widget will show which Communities the user is already following and which not, from a list predefined by the Client.

- PRECONFIGURATION IN DASHBOARD:

You can optionally choose the Likes in the Dashboard menu.

- OUTPUT:

An example of a call for the Engagement Widget through Xeerpa Connect Backend Auth PHPs looks like this:

An example of an URL through Xeerpa Connect Discovery including the call for the Engagement Widget looks like this:

Interests Widget:

 The Interests Widget will provide data about the (sub)categories a particular user has shown more interest on.

 You can customize the parameters of the call with the specific parameters of the Interests Widget:

- OUTPUT:

After he logs in, a JSON array will be generated with the following format:

 An example of a call for the Interests Widget through Xeerpa Connect Backend Auth PHPs looks like this:

An example of an URL through Xeerpa Connect Discovery including the call for the Interests Widget looks like this:

Social Groups Widget:

 The Social Groups Widget will provide aggregated data about each user's interests. A Social Group is a set of subcategories predefined by the Client.

 You can customize the parameters of the call with the specific parameters of the Social Groups Widget:

- PRECONFIGURATION IN DASHBOARD:

The Client will set up the Social Groups in the Dashboard menu with 4 to 8 subcategories in each one.

- OUTPUT:

After the user logs in, a JSON array will be generated with the following format:

The levels can be A-E.

An example of a call for the Social Groups Widget through Xeerpa Connect Backend Auth PHPs looks like this:

An example of an URL through Xeerpa Connect Discovery including the call for the Social Groups Widget looks like this: