Dear José, thanks for asking this question, and for sharing the link to the Qlik video. Indeed, you are correct, that the way of navigating data, in order to explore it, that QlikView calls associative logic is similar to navigating data using Excel slicers, and having the aggregates calculated in real time by the xVelocity engine: both tools allow you the freedom not to have to follow an OLAP hierarchy and to be able to see, live, the underlying relationships between attributes, or dimension members related through facts, or dimension relationships in a snowflake-like schema. Before xVelocity (PowerPivot Vertipaq) arrived, Microsoft analytics relied on precalculated aggregates using MOLAP, (or live ROLAP) but with the restriction of a fixed hierarchy. This was very good for performance (MOLAP) and for having control over what aggregates should be available to query, but it was restrictive to the user, leading to a lack of agility of the solution. Bear in mind, however, that MOLAP can still outperform real-time, in-memory aggregations for certain types of complex data sets, especially in the financial world.

I am glad you asked about the term associative, as there are several concepts, which share a similar name. Associative Logic is a QlikView-patented technology. Association Rules is a generic, industry-standard data mining technique, that is often used for associative data modelling, which relies on a concept of itemsets (or baskets) and rules that describe, by means of probabilities, the likelihood of certain combinations of items existing in a basket, given other items’ presence. This technique, part of SQL Server Data Mining, can be used for Market Basket Analysis, demographic analysis, and trend analysis. It works by detecting which individual items are related to which other items by analysis of all itemsets. This can be a computationally complex task for larger data sets, and SQL Server comes with a Microsoft-patented AR Algorithm that does a good job of putting it within reach of many users, as well as it comes with a visualisation tool, that works in Excel and Visio.

On the other hand, slicers in Excel, or associative logic in QlikView require you to already have known relationships between items. Those have to be present in the underlying data, by means of rows that relate the items to each other, or by means of key-based relationships between tables. The tools do a great job of helping the user navigate and explore such an existing set of relationships, with the magic of in-memory speed, live calculation of aggregates and, in the case of BISM, your own DAX function formulas. In other words, Association Rules discovers the patterns, while the  other two techniques help you explore already known patterns.