With our e-commerce data science
we support the research & development
of our AI technology

epoq Suite

Startseite » Personalization » Technology » Data Science

Our e-commerce data science
ensures the continuous development
of our AI technology

E-commerce data science is a data science that deals with the extraction of knowledge from data in the field of digital commerce. For this purpose, large amounts of data, also known as “big data”, are analysed and “smart data” extracted from it (data mining). This relevant data is finally used via artificial intelligence methods (machine learning) to derive recommendations for action for e-commerce. This mostly involves predicting purchasing behaviour (predictive analytics) with regard to certain personalisation measures. Therefore, our e-commerce data science is an important component of our e-commerce technology in order to continuously develop personalisation.

These data are relevant for our e-commerce data science

For our e-commerce data science, the click and purchase behaviour of your digital trade is essential, as well as your product catalogue. However, other data can also be included, e.g. from:

  • your CRM system,
  • your stationary trade (purchase data, customer cards, etc.)
  • various external data sources (e.g. returns data from a merchandise management system)

This is how the further development
of our e-commerce technology takes place

Digital Commerce

Analysis

Our Data Science team analyses the daily happenings in digital commerce and thinks its way into the business areas and products.

Development

Theories

Our Data Science team uses the accumulated knowledge from the field of consumer psychology to develop new theories that lead to even better personalisation.

Design

AI Engine

Our data science team then transfers the developed theories into the design of our AI engine. A specific goal is specified, which is to be optimised within a roughly defined framework. This allows us to quickly determine which theory is best and which parameter settings should be used.

A/B test

Verification

We use A/B tests to verify the effectiveness of the theory gained in relation to the status quo. Here we apply statistical measures to achieve a high degree of decision accuracy.

Platform

Recording

The tested theory becomes part of our personalisation platform and is available to you for the full personalisation of your online shop.

The essential processes for the
further development of our
AI engine

Data Mining

Data mining is about exploratory data analysis. We use statistical methods to gain valuable insights from the data. We use classical and Bayesian statistics for this purpose.

Classical Statistics

This is a quantitative data evaluation. It analyses the occurrence of a variable under certain conditions, such as the average level of sales. We use this to test our theories within the A/B tests for coincidence or patterns.

Bayesian Statistics

Here, the probability between two elements is calculated, such as the probability that it will rain if a black cloud occurs (calculation of composite probabilities). In other words, if someone buys product A, how likely is it that that person will buy product B? This is how we test our theories within the A/B tests for coincidence or pattern.

Machine Learning

Artificial intelligence is at home in the field of machine learning. Methods and techniques from supervised and reinforcement learning are used to let the AI engine learn independently.

Supervised Learning

Here, an input variable and a target variable are defined. The algorithm must move from the input variable to the target variable. It is given examples to learn how to achieve this target.

Reinforcement Learning

Learning by doing with subsequent adaptation based on the result. This is how the algorithm learns what works best.

The right use of
data mining and machine learning
is what counts

Our e-commerce data science uses predictive analytics to predict future events from historical data. For this purpose, a mathematical model is created that captures important trends. The right combination of data mining and machine learning techniques, combined with a high degree of consumer psychology, enable our data science team to make valuable predictions for the future via predictive analytics. In this way, our data science team is able to continuously develop our AI engine.

You don’t use personalisation in your online shop yet?

Find out more about the personalisation options now!

Learn more about personalisation now

The latest blog articles
on AI technology

Blog Post

The Reinforcement Learning Process: How To Use Reinforcement Learning to Increase the Profitability of Your Online Shop

Reinforcement learning is a type artificial intelligence, where a so-called agent learns to interact with its environment in the best way possible. In recent years, various applications have emerged which have made the process popular. Programs have been developed that can beat humans at games like chess or go, or even for simple Atari games. These programs help robots successfully play football or perform daring acrobatic helicopter flights. In this three-part blog series, we’re going to show you how the reinforcement learning process can be used to personalise online store recommendations.

Eric Mende: 01. Feb 2018

Learn More
All articles