Compliance dossier — training data (AI Act, Article 10)

Automatically generated skeleton from the dataset analysis. The "to be completed by the expert" blocks require human review (provenance, compliance judgment, mitigation measures).

1. Identity & purpose

System / model: _____
Dossier version: _____
Date: _____
Dataset: src
Intended purpose: _____

_To be completed by the expert — intended purpose and what the data is meant to represent_

2. Provenance & legal basis

Origin of each source, licences/contracts, and — for personal data — initial purpose and GDPR legal basis.

⚠️ Potential personal data detected: name, first, last, dob — a GDPR legal basis is required for these columns.

_To be completed by the expert — provenance per source (separate flow §16), licences, GDPR legal basis_

3. Composition _(filled automatically)_

Volume: 7214 rows, 53 columns.
Variables:

Column	Type
`id`	BIGINT
`name`	VARCHAR
`first`	VARCHAR
`last`	VARCHAR
`compas_screening_date`	DATE
`sex`	VARCHAR
`dob`	DATE
`age`	BIGINT
`age_cat`	VARCHAR
`race`	VARCHAR
`juv_fel_count`	BIGINT
`decile_score`	BIGINT
`juv_misd_count`	BIGINT
`juv_other_count`	BIGINT
`priors_count`	BIGINT
`days_b_screening_arrest`	BIGINT
`c_jail_in`	TIMESTAMP
`c_jail_out`	TIMESTAMP
`c_case_number`	VARCHAR
`c_offense_date`	DATE
`c_arrest_date`	DATE
`c_days_from_compas`	BIGINT
`c_charge_degree`	VARCHAR
`c_charge_desc`	VARCHAR
`is_recid`	BIGINT
`r_case_number`	VARCHAR
`r_charge_degree`	VARCHAR
`r_days_from_arrest`	BIGINT
`r_offense_date`	DATE
`r_charge_desc`	VARCHAR
`r_jail_in`	DATE
`r_jail_out`	DATE
`violent_recid`	VARCHAR
`is_violent_recid`	BIGINT
`vr_case_number`	VARCHAR
`vr_charge_degree`	VARCHAR
`vr_offense_date`	DATE
`vr_charge_desc`	VARCHAR
`type_of_assessment`	VARCHAR
`decile_score_1`	BIGINT
`score_text`	VARCHAR
`screening_date`	DATE
`v_type_of_assessment`	VARCHAR
`v_decile_score`	BIGINT
`v_score_text`	VARCHAR
`v_screening_date`	DATE
`in_custody`	DATE
`out_custody`	DATE
`priors_count_1`	BIGINT
`start`	BIGINT
`end`	BIGINT
`event`	BIGINT
`two_year_recid`	BIGINT

Declared sensitive attributes: race, sex, age_cat.

Detected PII columns: name, first, last, dob.

_To be completed by the expert — geographic / contextual / behavioural scope_

4. Preparation _(automatic observations)_

Missing values: days_b_screening_arrest (307) → handling strategy to be documented.

_To be completed by the expert — transformation log: collection, cleaning, labelling, enrichment, aggregation_

5. Quality _(filled automatically)_

Completeness: 307 missing value(s) (days_b_screening_arrest).
Accuracy / outliers: age (1).

_To be completed by the expert — representativeness and relevance to the purpose_

6. Bias

Automatic analysis

race — disparity of the « score_text » rate: ratio 4.83 ⚠️

Group	Count	« score_text » rate
`Native American`	18	33.3 %
`African-American`	3696	27.7 %
`Caucasian`	2454	11.2 %
`Hispanic`	637	10.5 %
`Asian`	32	9.4 %
`Other`	377	6.9 %

sex — disparity of the « score_text » rate: ratio 1.53 ⚠️

Group	Count	« score_text » rate
`Male`	5819	20.8 %
`Female`	1395	13.6 %

age_cat — disparity of the « score_text » rate: ratio 3.65 ⚠️

Group	Count	« score_text » rate
`Less than 25`	1529	29.6 %
`25 - 45`	4109	20.0 %
`Greater than 45`	1576	8.1 %

Judgment & measures

_To be completed by the expert — impact on health/safety/fundamental rights and detection/prevention/mitigation measures_

7. Gaps & limitations

Detected points to examine as potential gaps:

[low] missing — days_b_screening_arrest
[low] missing — c_jail_in
[low] missing — c_jail_out
[low] missing — c_case_number
[low] missing — c_offense_date
[high] missing — c_arrest_date
[low] missing — c_days_from_compas
[low] missing — c_charge_desc
[high] missing — r_case_number
[high] missing — r_charge_degree
[high] missing — r_days_from_arrest
[high] missing — r_offense_date
[high] missing — r_charge_desc
[high] missing — r_jail_in
[high] missing — r_jail_out
[high] missing — violent_recid
[high] missing — vr_case_number
[high] missing — vr_charge_degree
[high] missing — vr_offense_date
[high] missing — vr_charge_desc
[low] missing — in_custody
[low] missing — out_custody
[low] outliers — age
[low] outliers — priors_count
[low] outliers — days_b_screening_arrest
[low] outliers — c_days_from_compas
[low] outliers — r_days_from_arrest
[low] outliers — priors_count_1
[low] outliers — start
[high] pii — name
[high] pii — first
[high] pii — last
[high] pii — dob
[high] bias — race
[medium] bias — sex
[high] bias — age_cat

_To be completed by the expert — identified and addressed gaps; out-of-scope uses_

8. Governance & traceability

Analysis tool version: 0.2.0

_To be completed by the expert — responsibilities, audit log, versioning, maintenance/updates_