The 3 Flavors of Sitecore Cortex: How Machine Learning Helps Marketing

Most marketers today face a huge challenge with personalised marketing, more so about managing it at scale. To effectively set up personalization rules includes data analysis, hypotheses, content curation and manual configuration. Sometimes, it does become a little challenging to manage multiple rules across multiple pages for multiple scenarios and also figure what’s working or not.

With machine learning, this task becomes much simpler since no guess work is required. Using this model, a brand can analyse successful behaviour to send out relevant content and adopt targeted messaging on e-commerce sites, based on a customer’s journey.

Machine Learning 101

a. Who is it for

b. When it makes sense

c. Main types of ML, main use cases

d. Main elements and the process of ML

Lets dig a bit into it. Every machine learning algorithm has three components:

Representation: how to represent knowledge. Examples include decision trees, sets of rules, instances, graphical models, neural networks, support vector machines, model ensembles and others.

Evaluation: the way to evaluate candidate programs (hypotheses). Examples include accuracy, prediction and recall, squared error, likelihood, posterior probability, cost, margin, entropy k-L divergence and others.

Optimization: the way candidate programs are generated known as the search process. For example combinatorial optimization, convex optimization, constrained optimization. All machine learning algorithms are combinations of these three components.

Machine learning is again classified as - Supervised learning, Unsupervised learning and Reinforcement learning

Lets talk more about these with use cases.

Supervised learning is task-oriented. It is highly focused on a singular task, feeding more and more examples to the algorithm until it can accurately perform on that task. This is the learning type that we will most likely encounter. Face Recognition is one fine example.

Facebook uses the supervised learning that is trained to recognize your face. Having a system that takes a photo, finds faces, and guesses who that is in the photo (suggesting a tag) is a supervised process. Spam Classification and Advertisement Popularity are few other cases where supervised learning is used.

Unsupervised learning is very much the opposite of supervised learning. It features no labels. Instead, our algorithm would be fed a lot of data and given the tools to understand the properties of the data. From there, it can learn to group, cluster, and/or organize the data in a way such that a human (or other intelligent algorithm) can come in and make sense of the newly organized data.

YouTube or Netflix uses Unsupervised learning. They are developed with algorithms to know things about videos, their length, their genre, user details etc. and then we are recommendations with videos of similar genre.

Reinforcement learning is fairly different when compared to supervised and unsupervised learning where we can easily see the relationship between supervised and unsupervised (the presence or absence of labels), the relationship to reinforcement learning is a bit murkier.

One of the most common places to look at reinforcement learning is in learning to play games. Look at Googleís reinforcement learning application, AlphaZero and AlphaGo which learned to play the game Go. Reinforcement learning is good for navigating complex environments.

So in short ML process involves:

-Identifying business goal (domain expertise is mandatory)

-Assess the data

-Modelling (R & CNTK to produce models, know more about R & CNTK in the coming sections of this article)

-Application(using cortex APIs in developing the required ML models)

Welcome Cortex

a. What is available with SXP 9.1 for a digital marketer How to get started and how to be successful with Cortex

b. Prerequisites for ML ñ strategy, activities and resources Data as an integral part of Cortex

c. Altudo Connectors Example of a successful ML program

d. Altudo case study

Sitecore 9.1 delivers omnichannel marketing at scale, natively integrated data insights, and enhanced behavioral tracking capabilities. Additional enhancements include Federated Authentication, WCAG 2.0 compliance in SXA, external triggers for Data Exchange Framework 2.1, as well as performance improvements for deployments. Please refer to release notes for all innovations and enhancements in Sitecore 9.1.