How does Bedrock's learning algorithm work?

Bedrock’s curriculum is supported by a deep learning algorithm, which has been designed to ensure continual development for each learner. It’s the magic behind the platform! The algorithm provides a personalised learning journey for every learner and encourages long term memory retention and mastery of the content.

What is the algorithm?

Bedrock’s algorithm is based upon three key principles: assess, teach, and reteach. This ensures newly acquired knowledge is re-served to learners at regular intervals, which is a crucial step in long term memory retention, and supports mastery of the content. Personal to each learner, this approach provides multiple exposures to the content that they have found most challenging in order to support them further.

What is recapping?

Recapping is when Bedrock serves a learner content they have seen before and have gotten correct. They must get it correct again in order for it to leave their recapping queue, and for the content to be considered ‘learnt.’ Recapping activities are served in the same order they were served to the learner originally, and reinforce understanding. In Bedrock Mapper, recapping is also called ‘memory checking.’

What is reteaching?

Reteaching serves a learner content that they have gotten incorrect during a lesson’s pre-test, post-test, or previous reteaching and recapping activities. Once a learner gets a word correct in reteaching, it will move to recapping.

How are recapping and reteaching served to the learner in Bedrock’s core curriculum?

At the start of each new lesson, reteaching and recapping are served to the learner in the following order:

• one reteaching activity
• one reteaching activity
• one recapping activity

If there are no words or Grammar knowledge items to recap or reteach, it is skipped until next time. For example, the learner could be served:

• one reteaching activity
• no reteaching activities
• one recapping activity

Knowledge organiser

Bedrock's learning algorithm also determines how a learner’s knowledge organiser is populated. A word will fall into the ‘Vocabulary I am learning’ or ‘Vocabulary I have learned’ columns depending on whether a learner has recapped it successfully or it is still part of their reteaching queue.

How is subject-specific recapping and reteaching served to the learner in Bedrock Mapper?

Reteaching is served slightly differently in Mapper. A learner’s pre-test and post-test answers inform the algorithm; if they get a word incorrect during the pre-test, it will be added to the reteaching queue regardless of their post-test answer.

In line with cognitive science, reteaching is served at the start of a new Mapper session, after a spaced amount of time from the initial exposure. The session will include 5 words from across the curriculum. Learners are served the explicit teaching content again, and 1 learning activity for each word (either the image activity, or the similar/different question.)

Once the learner has completed the reteaching session, they are returned to their subject-specific vocabulary lesson and served new content.

Alongside reteaching, Mapper also serves memory checks which assess whether the learner has ‘learned’ the word that they previously answered incorrectly. If they answer correctly, the word is moved from ‘learning’ to ‘learned.’ These memory check sessions also assess whether the learner has retained their knowledge of the word, in order to keep it in the ‘learned’ category.