Datasets:

GAIMHE
/

MIAAM-V2

license: cc-by-nc-sa-4.0
language:
  - fr
tags:
  - education
  - knowledge-tracing
  - student-modeling
  - mathematics
  - cognitive-science
task_categories:
  - other
size_categories:
  - 1M<n<10M
modality:
  - tabular
  - text
  - image
extra_gated_heading: Request access to this dataset
extra_gated_description: >-
  This dataset is released for research purposes only under CC-BY-NC-SA 4.0. Our
  team will review your request and may take a few days to process it.
extra_gated_button_content: Submit request
extra_gated_prompt: >-
  This dataset contains anonymized interaction data from primary and secondary
  school students in France and Côte d’Ivoire, including minors.  It is released
  strictly for non-commercial research purposes. Users must not attempt to
  re-identify any individual in the dataset, and must report any personally
  identifiable information (PII) encountered to the dataset authors. By
  submitting this form, you confirm that you have read and agree to the
  CC-BY-NC-SA 4.0 license terms.
extra_gated_fields:
  Name: text
  Affiliation type:
    type: select
    options:
      - University / Academic lab
      - Industry
      - label: Other
        value: other
  Affiliation name: text
  Affiliation website: text
  Country: country
  Email address: text
  Personal webpage: text
  Intended use of the dataset:
    type: select
    options:
      - Knowledge Tracing research
      - Student modeling research
      - LLM / VLM research
      - Educational data mining
      - label: Other
        value: other
  I agree to use this dataset for non-commercial purposes ONLY: checkbox
  I agree to the license terms: checkbox
  I agree not to attempt to re-identify any student or individual present in this dataset: checkbox
  I agree to report any personally identifiable information (PII) I may encounter to the dataset authors: checkbox
configs:
  - config_name: raw interactions
    default: true
    data_files:
      - split: train
        path: data/maths_data.parquet
  - config_name: filtered interactions
    data_files:
      - split: train
        path: data/maths_data_filtered.parquet
  - config_name: exercises_table
    data_files:
      - split: train
        path: data/maths_exercises_table.parquet

MIAAM V2.1: Mathematics Interaction Dataset

Overview

MIAAM is a dataset of student interactions with mathematics exercises from three digital learning platforms used in France and Côte d’Ivoire. It covers early primary school, middle school, and high-school remediation, with exercises on topics such as numbers, arithmetic problems, geometry, fractions, and data.

The dataset brings together three sources:

Adaptiv’Math (am): early primary mathematics, including numbers and arithmetic problem solving.
Adaptiv College (adaptiv_college): geometry, proportions and fractions, and data organization.
Adaptiv World (adaptiv_world): number sense, data organization, and functions. These data were collected in Côte d’Ivoire.

Each platform contains more material than the modules included in this release. The included modules are listed below.

The exercises are organized as modules → objectives → activities → exercises. An exercise can appear in more than one activity or module.

You can explore the dataset with the MIAAM visualization tool.

Included modules

Source	Code	Module
Adaptiv’Math	M1	Nombres et calcul
Adaptiv’Math	M31	Résolution de problèmes arithmétiques - Niveau 1
Adaptiv’Math	M32	Résolution de problèmes arithmétiques - Niveau 2
Adaptiv’Math	M33	Résolution de problèmes arithmétiques - Niveau 3
Adaptiv College	M101	Espace et Géométrie
Adaptiv College	M102	Proportions et fractions
Adaptiv College	M103	Organisation et gestion des données
Adaptiv World	M101	Réapprentissage du sens des nombres
Adaptiv World	M103	Organisation et gestion de données, fonctions

These module codes are labels. The data files use UUIDs for module_id, and the same label can refer to different modules in different sources.

Work modes

The work_mode column indicates how an exercise was assigned:

adaptive-test: an adaptive placement test estimates a student’s starting level in a module.
playlist: a teacher assigns a predefined set of exercises.
zpdes: an adaptive system recommends work based on the student’s estimated level and the relationships between activities. ZPDES stands for Zone of Proximal Development and Empirical Success.

All three sources contain adaptive-test and zpdes interactions. The playlist mode appears only in Adaptiv World and contains four attempts by one student.

Data files

The Parquet and JSON files are in the data/ directory. Each row in an interaction table represents one attempt at an exercise. A student may attempt the same exercise more than once.

`maths_data.parquet`

This is the main interaction table. It contains 4,309,251 attempts, from 26,383 students, on 8,831 exercises. Student and exercise counts distinguish identifiers from different sources.

Source	Students	Attempts	Exercises
Adaptiv’Math	24,439	4,109,574	5,447
Adaptiv College	867	55,074	2,029
Adaptiv World	1,077	144,603	1,355

The table has 14 columns:

Column	Description
`user_id`	Student identifier. Use with `source`.
`classroom_id`	Classroom identifier. Missing for 1,018 attempts in the main table.
`playlist_or_module_id`	Assignment context. In this release, it matches a catalogue `module_id`, including for playlist attempts.
`objective_id`	Learning objective associated with the attempt.
`activity_id`	Activity associated with the attempt.
`exercise_id`	Exercise identifier. Use with `source`.
`created_at`	Shifted timestamp; see Timestamps.
`data_correct`	Boolean indicating whether the response was correct.
`work_mode`	How the exercise was assigned: `adaptive-test`, `playlist`, or `zpdes`.
`data_answer`	Recorded response as a string. Its format depends on the exercise’s gameplay type.
`data_duration`	Recorded response duration, in its original units. The units are not specified in this documentation.
`source`	Dataset of origin: `am`, `adaptiv_college`, or `adaptiv_world`.
`attempt_index`	Number of attempts by the student on this exercise, starting at 1. Indices can have gaps in the filtered table.
`session_id`	Session identifier derived from login time. Missing for 100,983 attempts in the main table.

Duration values: The main table contains nine negative values and some very large values. Check this column before using it in an analysis.

Timestamps

The dates in created_at are artificial. For each student, the earliest available login or attempt timestamp after the initial filtering is shifted to 2000-01-01 00:00:00. The elapsed time between that reference and each attempt is preserved, including across sessions and days.

The same reference is used in both interaction tables, so the first attempt in either table does not necessarily start at zero. Timestamps have no timezone information and should not be compared as real calendar dates across students. To order a student’s attempts, sort by source, user_id, and created_at; attempts can have identical timestamps.

`maths_data_filtered.parquet`

This table has the same columns as the main table, with additional filtering. It removes final placement-test blocks with fewer than five attempts, excludes students whose remaining history consists only of placement tests, and applies further filters, including a minimum of five remaining attempts per student. The preprocessing notebook documents the full procedure.

It contains 4,277,081 attempts, from 22,237 students, on 8,803 exercises.

Source	Students	Attempts	Exercises
Adaptiv’Math	20,823	4,082,192	5,421
Adaptiv College	557	52,046	2,027
Adaptiv World	857	142,843	1,355

`maths_exercises_table.parquet`

This table describes the exercises and where they appear in the learning content. It contains 9,167 exercises and 9,938 rows. Some exercises have more than one row because they appear in multiple activities or modules. The catalogue also includes exercises with no attempts in the interaction table.

Column	Description
`exercise_id`	Exercise identifier. Use with `source`.
`gameplay_type`	Interaction format, such as `MULTIPLE_CHOICE` or `INPUT_FIELD`.
`content`	Exercise content, including instructions, questions, answers, feedback, and image references where available.
`module_id`	Identifier of the module.
`module_name`	Name of the module.
`objective_id`	Identifier of the objective.
`objective_name`	Name of the objective.
`objective_pedagogical_intent`	Intended learning goal of the objective.
`activity_id`	Identifier of the activity.
`activity_name`	Name of the activity.
`activity_pedagogical_intent`	Intended learning goal of the activity.
`source`	Dataset of origin.

Most content values use JSON, but some Adaptiv’Math values use Python dictionary syntax. There are 26 gameplay types in the catalogue.

`maths_dependencies.json`

This file describes the structure and prerequisites of the included modules. For each activity, it provides its code, titles, exercises, prerequisite activities, and activities it unlocks.

The file contains 9 modules, 102 objective entries, 553 activity entries, and 586 prerequisite links. Objective and activity entries are counted each time they appear in the hierarchy.

`screenshots/`

Each exercise in the catalogue has a screenshot in two sizes:

screenshots/raw/<source>/<exercise_id>.png: original exported resolution.
screenshots/compressed/<source>/<exercise_id>.png: resized to 960 pixels wide, preserving the aspect ratio.

There are 9,167 screenshots of each size. Include source in the path because exercise IDs can be shared by different sources.

`descriptions.json`

This file contains a text description for each exercise in the catalogue. Use the key <source>/<exercise_id> to retrieve a description. The file also contains older entries keyed by exercise_id alone; prefer the source-specific keys.

Working with the data

Identifiers and joins

Use source together with user_id to identify a student, and together with exercise_id to identify an exercise. An ID on its own is not necessarily unique across sources.

To match an interaction to its exercise metadata in the relevant activity, join on:

source, objective_id, activity_id, exercise_id

These columns identify a row in the exercise table in this release. Joining only on source and exercise_id can duplicate attempts when an exercise appears in more than one activity or module.

Reading the main table

For example, with Python and Polars:

import polars as pl

# Each row represents one attempt at an exercise.
interactions = pl.read_parquet("data/maths_data.parquet")
print(interactions.head())

Reproducing the filtered table

The preprocessing notebook rebuilds the filtered table from the main interaction table and exercise catalogue. It requires Python and Polars.

Visualization tool

The MIAAM visualization tool lets you explore dataset statistics, student trajectories, work modes, exercise formats, and the organization of the modules and activities.

License

Dataset: CC BY-NC-SA 4.0 — non-commercial use, with attribution and share-alike requirements.
Preprocessing scripts: MIT License.

Release

V2.1 — 2026-10-28.