bihungba1101/slam-es-en-knowledge-tracing

SLAM Spanish-English Knowledge Tracing

This repository contains a Hugging Face-compatible conversion of the Spanish-English track from the 2018 Second Language Acquisition Modeling (SLAM) shared task.

Each row represents one learner exercise attempt. The original token-level response labels are retained in token_labels. The correct field is strict whole-attempt correctness: it is 1 only when every response token is correct.

This dataset is suitable for longitudinal knowledge tracing because it records repeated interactions for anonymous learners, relative elapsed time, response correctness, challenge type, session type, and response time. Challenge types are not restricted to multiple-choice questions; they include language-learning activities such as listening, reverse tapping, and reverse translation.

Dataset schema

The following example shows one converted interaction:

{
  "user_id": "+H9QWAV4",
  "interaction_index": 4,
  "exercise_id": "Ii4fn+U101",
  "prompt": "The knife is black.",
  "country": "CA",
  "client": "ios",
  "session": "lesson",
  "format": "reverse_translate",
  "days": 13.424,
  "response_time_seconds": 25.0,
  "token_ids": [
    "Ii4fn+U10101",
    "Ii4fn+U10102",
    "Ii4fn+U10103",
    "Ii4fn+U10104"
  ],
  "response_tokens": ["El", "cuchillo", "es", "negro"],
  "pos_tags": ["DET", "NOUN", "VERB", "ADJ"],
  "morphological_features": [
    "Definite=Def|Gender=Masc|Number=Sing|PronType=Art|fPOS=DET++",
    "Gender=Masc|Number=Sing|fPOS=NOUN++",
    "Mood=Ind|Number=Sing|Person=3|Tense=Pres|VerbForm=Fin|fPOS=VERB++",
    "Gender=Masc|Number=Sing|fPOS=ADJ++"
  ],
  "dependency_relations": ["det", "nsubj", "cop", "ROOT"],
  "dependency_heads": [2, 4, 4, 0],
  "token_labels": [1, 1, 0, 1],
  "token_accuracy": 0.75,
  "correct": 0,
  "prompt_available": True
}

Each token-level list is positionally aligned. For example, response_tokens[2] is "es", its corresponding token_ids[2] is "Ii4fn+U10103", its pos_tags[2] value is "VERB", its dependency_relations[2] value is "cop", and its token_labels[2] value is 0.

The linguistic annotations are Universal Dependencies-style annotations from the original SLAM files. They describe the learner's response rather than the prompt. For example, DET means determiner, NOUN means noun, and ROOT identifies the syntactic root. A dependency head is a one-based token position; 0 means that the token is the root.

Feature reference

Feature	Type	Description	Example
user_id	string	Anonymous learner identifier. Rows with the same value belong to the same learner.	"+H9QWAV4"
interaction_index	integer	Sequential attempt number for that learner within the source split.	4
exercise_id	string	Identifier derived from the response token IDs for this attempt. It identifies an attempt instance and should not automatically be treated as a stable question/content ID.	"Ii4fn+U101"
prompt	string or null	English source prompt when the original interaction provides one. It is often absent for listening activities.	"The knife is black."
country	string	Country metadata associated with the learner.	"CA"
client	string	Application or device platform used for the interaction.	"ios"
session	string	Learning session type, such as lesson or practice.	"lesson"
format	string	Challenge type, such as listen, reverse_tap, or reverse_translate.	"reverse_translate"
days	float	Learner-relative elapsed time in days from the beginning of the learner's observed activity timeline. It is not a calendar date.	13.424
response_time_seconds	float or null	Time spent on the attempt in seconds. Missing source values are null.	25.0
token_ids	list of strings	Source identifiers for the response tokens.	["Ii4fn+U10101", "Ii4fn+U10102"]
response_tokens	list of strings	Tokens produced or selected by the learner.	["El", "cuchillo", "es", "negro"]
pos_tags	list of strings	Universal part-of-speech tags aligned with response_tokens.	["DET", "NOUN", "VERB", "ADJ"]
morphological_features	list of strings	Morphological annotations aligned with response_tokens, including features such as gender, number, tense, person, and fine-grained POS.	`["Gender=Masc
dependency_relations	list of strings	Syntactic dependency relation for each response token, such as det, nsubj, cop, or ROOT.	["det", "nsubj", "cop", "ROOT"]
dependency_heads	list of integers	One-based syntactic head position for each response token; 0 identifies the root.	[2, 4, 4, 0]
token_labels	list of integers	Token correctness labels aligned with response_tokens; 1 means correct and 0 means incorrect.	[1, 1, 0, 1]
token_accuracy	float	Fraction of response tokens marked correct.	0.75
correct	integer	Strict whole-attempt target; 1 only when every token is correct, otherwise 0.	0
prompt_available	boolean	Whether a prompt is available for this interaction.	True

Understanding time and learner histories

The days feature is learner-relative elapsed time. It indicates how many days had passed since the beginning of that learner's recorded activity timeline. It is not an account creation date, a calendar date, or a global timestamp shared by all learners.

For example, the following values indicate activity near the beginning of a learner timeline, followed by activity roughly 13 days later:

days:0.009
days:13.424
days:14.710

The time gap between consecutive interactions is calculated as:

time_gap_days = days_current - days_previous

For example, 14.710 - 13.424 = 1.286 days, or approximately 30.86 hours. This supports modelling practice spacing, forgetting effects, and learning progression.

The first interaction visible in a particular split does not necessarily have days = 0. A learner may have earlier interactions in another source split, so a validation or test record can begin at a later relative time.

Correctness and modelling targets

The source data provides correctness at the response-token level. The conversion preserves these labels and derives two summary targets:

• token_accuracy: the fraction of correct response tokens.
• correct: strict interaction-level correctness.

The strict interaction label is defined as:

correct = 1 if every token label is 1; otherwise correct = 0

For example:

token_labels	token_accuracy	correct
[1, 1, 1]	1.00	1
[1, 1, 0, 1]	0.75	0
[0, 0]	0.00	0

This supports two related tasks:

1. Interaction-level knowledge tracing: predict whether the learner will complete the next exercise correctly.
2. Token-level knowledge tracing: predict whether each response token will be correct.

A typical interaction-level KT objective is:

P(correct[t+1] = 1 | learner history through interaction t, target exercise[t+1])

The most important fields for this task are user_id, days, interaction_index, format, prompt, correct, response_time_seconds, and session.

Example learner sequence

Rows can be grouped by user_id and ordered by days and interaction_index:

interaction_index	days	format	prompt	correct
1	0.009	listen	null	0
2	0.009	listen	null	1
3	0.009	reverse_translate	The bread	1
4	13.424	reverse_translate	The knife is black.	0
5	13.424	reverse_tap	I have plates.	1

A knowledge-tracing model can use interactions 1 through 4 to predict the outcome of interaction 5.

Linguistic annotation example

The original files contain rows such as:

Lmzbsr1h0201  El       DET   Definite=Def|Gender=Masc|Number=Sing|PronType=Art|fPOS=DET++  det   2
Lmzbsr1h0202  espejo   NOUN  Gender=Masc|Number=Sing|fPOS=NOUN++                       ROOT  0

In the converted dataset, these rows are stored inside one interaction as aligned lists:

{
  "exercise_id": "Lmzbsr1h02",
  "response_tokens": ["El", "espejo"],
  "pos_tags": ["DET", "NOUN"],
  "morphological_features": [
    "Definite=Def|Gender=Masc|Number=Sing|PronType=Art|fPOS=DET++",
    "Gender=Masc|Number=Sing|fPOS=NOUN++"
  ],
  "dependency_relations": ["det", "ROOT"],
  "dependency_heads": [2, 0],
  "token_labels": [0, 0]
}

Here, El is a masculine singular definite determiner attached to token position 2, while espejo is a masculine singular noun and the syntactic root. The correctness labels shown above are illustrative of the converted row; use the published row values for the exact labels in a particular split.

Splits

• train: source training split.
• validation: source development split.
• test: source test split.

The converted split sizes are:

Split	Interactions
train	731,896
validation	96,003
test	93,145

The source files contain approximately 2,640 anonymous learners per split. The same learner IDs occur across the supplied splits, so these splits support longitudinal evaluation but are not learner-disjoint cold-start splits. Create new user-disjoint splits when evaluating generalization to unseen learners.

Load

from datasets import load_dataset

dataset = load_dataset("bihungba1101/slam-es-en-knowledge-tracing")

Important: the supplied source splits reuse learner IDs across splits. They are not learner-disjoint cold-start splits.

Citation and license

See CITING.md and LICENSE.md. This conversion does not change the source dataset license or citation requirements.

Conversion

The Parquet files were generated by convert_to_hf.py in this repository. The converter groups source token rows into one interaction row, joins development and test labels from their .key files, preserves token-level correctness and all linguistic annotations, and derives strict interaction-level correctness.