12 November 2015

Recommendation, personalisation en persuasion

In e-commerce, but also in other sectors, recommendations, personalisation and persuasion are important tools to influence transactions (customer decisions). Anchormen delivers a smart combination of (big) data technology and machine/deep learning applications with which all the available data contribute to optimisation of the decision-making process.

Afficial intelligence, machine learning in unstructured data

Both streaming and batch data containing images, videos, text or other unstructured data that can be analysed require smart software. We implement streaming platforms such as Spark or Storm for this. For batch data we for example make use of the Hadoop batch platform. We combine this with smart components and algorithms from the machine/deep learning.

We proceed as follows:

  • Set up a product management system (PDM) and develop a taxonomy: a structure that enables content about for example products, customers, events and organisation to be clustered in groups so that they can be easily identified, examined and retrieved.
  • Churn: the challenge for Data Science is not in implementing an algorithm, but in getting the right algorithms to find an answer to your request in real time. In this case, the question is: how many customers are going to leave my customer base? In other words: what is my churn rate? Read more here.
  • Deep Learning Profiling