SentimentForecastStored public headlines

Brand news

Recent source-linked headlines from GDELT and Google News RSS. Only clearly relevant matches appear in the gauge and article list.

1272 of 1518 stored brand–article matches for automotive qualify; 246 are uncertain and 0 are excluded quote pages. Counts reflect sampled tracked coverage, not all published news.

Headline sentiment · experimental

+6

Leans positive

−100 more negative0+100 more positive

Based on 85 scored headlines for all 6 automotive brands: 15 positive, 60 neutral, 10 negative, 3 unscored (not English or mixed) · Oct 13, 2026 – Oct 17, 2026. Same articles as the list below. Experimental word-based headline tone only—not consumer sentiment or a forecast.

88 articles · all 6 automotive brands

Uncertain matches246 uncertain · 0 excluded

The feeds returned these for the brand, but the headline alone does not clearly show it is about the brand. Up to 15 per brand are shown (deduplicated across brands). They are shown for transparency, carry no tone score, and are not counted in the gauge or article totals. 0 stock quote/forecast pages excluded entirely.

Source freshness by brand12 source checks · 12 need attention

From the collection ledger (records begin Sep 27, 2026 06:29 UTC; earlier runs were not logged). Freshness is the last time a fetch was actually saved — not the newest article date. Each source is shown separately.

BrandSourceLatest attemptOutcomeLast successful save
ToyotaGoogle News RSSNov 2, 2026 03:18 UTCSkippedSkipped: this run's time budget for the provider was used upNo logged success · source-wide Oct 17, 2026 06:35 UTC (pre-ledger)
ToyotaGDELTNov 2, 2026 03:19 UTCSkippedSkipped: provider was failing repeatedly in this runNo logged success · source-wide Nov 2, 2026 03:19 UTC (pre-ledger)
FordGoogle News RSSNov 2, 2026 03:18 UTCSkippedSkipped: this run's time budget for the provider was used upNo logged success · source-wide Oct 17, 2026 06:35 UTC (pre-ledger)
FordGDELTNov 2, 2026 03:19 UTCSkippedSkipped: provider was failing repeatedly in this runNo logged success · source-wide Nov 2, 2026 03:19 UTC (pre-ledger)
HondaGoogle News RSSNov 2, 2026 03:18 UTCSkippedSkipped: this run's time budget for the provider was used upNo logged success · source-wide Oct 17, 2026 06:35 UTC (pre-ledger)
HondaGDELTNov 2, 2026 03:19 UTCSkippedSkipped: provider was failing repeatedly in this runNo logged success · source-wide Nov 2, 2026 03:19 UTC (pre-ledger)
ChevroletGoogle News RSSNov 2, 2026 03:18 UTCSkippedSkipped: this run's time budget for the provider was used upNo logged success · source-wide Oct 17, 2026 06:35 UTC (pre-ledger)
ChevroletGDELTNov 2, 2026 03:19 UTCSkippedSkipped: provider was failing repeatedly in this runNo logged success · source-wide Nov 2, 2026 03:19 UTC (pre-ledger)
TeslaGoogle News RSSNov 2, 2026 03:18 UTCSkippedSkipped: this run's time budget for the provider was used upNo logged success · source-wide Oct 17, 2026 06:35 UTC (pre-ledger)
TeslaGDELTNov 2, 2026 03:19 UTCSkippedSkipped: provider was failing repeatedly in this runNo logged success · source-wide Nov 2, 2026 03:19 UTC (pre-ledger)
BMWGoogle News RSSNov 2, 2026 03:18 UTCSkippedSkipped: this run's time budget for the provider was used upNo logged success · source-wide Oct 17, 2026 06:35 UTC (pre-ledger)
BMWGDELTNov 2, 2026 03:19 UTCSkippedSkipped: provider was failing repeatedly in this runNo logged success · source-wide Nov 2, 2026 03:19 UTC (pre-ledger)
About this data Scoring, relevance and limits

How headline tone is calculated

Each headline (a trailing publisher suffix is removed only if it matches the stored publisher) is scored with the open-source sentiment library, which sums AFINN-165 English word scores (−5 to +5) and flips a word's score when it directly follows a negator such as "not". A total above 0 is positive, below 0 is negative, exactly 0 is neutral. Headlines not recognized as English are unscored and left out. Scoring version tone-v2 also neutralizes a few misleading words ("straight", "exclusive", "backs", "block", "super"), adds −2 for one obvious business-decline or production-halt phrase (e.g. "sales fall", "halts production"), and leaves mixed headlines (both directions, net within ±1) unscored rather than guessing. These rules were checked against real stored headlines, not a labeled benchmark; the model is not validated.

Relevance first. Before sampling the latest 15 per brand, every stored headline is classified by fixed rules (headline and link only — articles are not read): clear when the brand or a recognized name is in the headline (for "New Balance" and "Puma", which are also ordinary words or surnames, sportswear or company context is also required); uncertain when the brand is missing from the headline or the name looks like a person, animal or team, or the name refers to another company such as Puma Energy; excluded only for evergreen stock-quote pages with an unmistakable quote-page template or a known quote publisher's page format. Earnings, forecast and profile journalism is kept. Only clear headlines count toward the list, gauge and dashboard article counts. Stored articles are never deleted. This is not verified full-article relevance.

Gauge = round(100 × (positive − negative) ÷ (positive + neutral + negative))

This is automated word-based tone of headlines only — not full-article analysis, brand-specific public opinion, consumer sentiment, sales or a forecast. It is not a validated model. Word lists miss sarcasm, context and ambiguity, and a headline may describe another company, athlete or event. Dates are publication dates for Google News and first-seen dates for GDELT. More about the project.