Deux articles, un même fait. Que disent-ils vraiment ?
https://infoscore.coThis text is a meta-description of the Info Score tool itself—a media literacy instrument designed for French high schools (AEFE network). It explains that the tool compares two articles on the same topic, extracts observable variables (sources, viewpoints, fact/commentary separation, emotional charge, uncertainty acknowledgment, context), and produces a letter grade from A to E using deterministic public rules. The text emphasizes pedagogical comparison and open questions for a 55-minute EMI session, noting that Info Score does not decide truth but makes differences visible.
Sub-scores
Representative quotes
- « Info Score ne décide pas de la vérité. »
- « Un outil pour la classe, pas un verdict. »
- « Une fonction déterministe applique ensuite les règles publiques (rules_v0.yaml) pour produire une note de A à E. »
Methodological notes
This text is itself a description of the Info Score methodology, not an article to be analyzed. It contains no journalistic content—no factual claims, sources, opinions, or emotional language. All counts related to claims, sources, opinion markers, and loaded language return zero. The text appropriately describes the tool's purpose and function in neutral, technical terms. The versioned rule reference (rules_v0.yaml) is verifiable.
Computation detail
Each step is reproducible from the extracted variables.
Show / hide full trace
Fact / commentary separation
- attribution_ratio = 0/1 = 0.000
- unmarked_density = 0/140 mots → 0.00/1000 → pénalité 0.000
- marker_bonus (transparence) = 0.000
- score = clamp(0.55·0.000 + 0.45·(1-0.000) + 0.000) = 0.450
Variables used
{
"claims_total": 1,
"claims_with_attribution": 0,
"opinion_markers_count": 0,
"unmarked_opinions_presented_as_facts": 0,
"word_count": 140
}
Source traceability
- named_density = 0/200 mots → score nommé 0.000
- non_anon_ratio = 0/1 = 0.500
- primary_ratio = 0/1 = 0.000
- links_bonus = 0 liens → 0.000
- score = clamp(0.35·0.000 + 0.30·0.500 + 0.25·0.000 + 0.000) = 0.150
Variables used
{
"external_links_count": 0,
"sources_anonymous_count": 0,
"sources_named_count": 0,
"sources_primary_count": 0
}
Plurality of viewpoints
- subject_type = factuel_incontesté (pluralité moins exigible)
- score = clamp(0.5 + 0·0.2) = 0.500
Variables used
{
"distinct_viewpoints_count": 0,
"subject_type": "factuel_incontest\u00e9"
}
Expression of uncertainty
- calibration = bien_calibré → 1.00
- hedging_present = False → bonus 0.00
- premature_conclusions = 0 → pénalité 0.000
- score = clamp(1.00 + 0.00 - 0.000) = 1.000
Variables used
{
"confidence_calibration": "bien_calibr\u00e9",
"hedging_present": false,
"premature_conclusions_count": 0
}
Tone and lexicon
- intensity = neutre → base 1.00
- loaded_density = 0.00/1000 mots → pénalité 0.000
- ad_hominem = 0 → pénalité 0.000
- rhetorical = 0 → pénalité 0.000
- sensational_headline = False → 0.00
- score = clamp(1.00 - 0.000 - 0.000 - 0.000 - 0.00) = 1.000
Variables used
{
"ad_hominem_attacks_count": 0,
"emotional_charge_intensity": "neutre",
"loaded_words_count": 0,
"rhetorical_questions_count": 0,
"sensationalist_headline": false
}
Contextualisation
- completeness = non_nécessaire → base 0.85
- perspectives_bonus (hist=False, géo=False, chiffres_verif=True) = 0.10
- missing_essential_info = 0 → pénalité 0.000
- score = clamp(0.85 + 0.10 - 0.000) = 0.950
Variables used
{
"dates_and_figures_verifiable": true,
"geographic_context_provided": false,
"historical_context_provided": false,
"missing_essential_info_count": 0,
"relevant_background_complete": "non_n\u00e9cessaire"
}