Content recommendations for increased relevance in digital media
Purpose and goal
Triggerbee offers a personalization service to e-retailers and media companies. Using behavioral data from the website in combination with customer data, Triggerbee´s customers can create relevance on the website. Through this project, we want to start applying machine learning to automatically identify different personas and categorize content. In this way, no predefined rules are needed for each customer, and we do not have to wait for different predefined conditions to be fulfilled to determine if a visitor belongs to a certain category, and what content to present.
Expected results and effects
The project aims to provide Triggerbee with the conditions to develop the basis for a commercial offering, which after the project has ended can be continuously enhanced together with Triggerbee´s customers and within the framework of Triggerbee´s own R&D budget. Through the project we will collaborate with reference customers where we aim to deliver measurably increased reading and conversion on their websites.
Planned approach and implementation
The project is divided into the following elements: * Evaluate technologies and available frameworks at our cloud provider * Learn to apply these technologies to analyze and detect patterns in large amounts of data * Develop application in the form of software "recommendation engine" that can be used by website owners to recommend content based on data The outcome can be compared with a control group on both participating customers´ websites and non-participating. In this way, we will be able to obtain statistically reliable data on how well the recommendation engine performs.