Stuff That Spins
SPIN Processed
Source arXiv Computation and Language export.arxiv.org Analyst
August 10, 2026 research research

ConstructCIE: A Dataset for Extracting Causal Information from Construction Accident Narratives

Positions the work as socially consequential by anchoring it to construction safety and OSHA data, implying public-good relevance without explicit virtue language.

View original on arxiv.org

Overview

Researchers released ConstructCIE, a manually annotated dataset for extracting hierarchical causal information from OSHA construction accident reports, revealing persistent gaps in LLM and sequence tagger performance on precise evidence-span extraction.

TL;DR

  • ConstructCIE is a new human-annotated dataset for causal information extraction from real construction accident reports.
  • It uses a hierarchical schema covering accident types, causal factors, sub-factors, and supporting evidence spans.
  • Evaluated models succeed at high-level accident classification but consistently fail at precise, span-level causal evidence extraction.

Key Stats

OSHA reports

source data

U.S. Occupational Safety and Health Administration incident narratives

Questions Answered

What happened?Who is involved?Why does this matter?

Keywords

causal information extractionconstruction safetyOSHAdatasetLLM evaluation

Narrative Frame

research framing

The Halo

Spin Score

35%

Emphasizes domain importance and manual curation; minimizes discussion of dataset scale, annotation consistency, or real-world deployment pathways.

What the story wants you to believe

That ConstructCIE is a credible, rigorously designed benchmark enabling meaningful evaluation of causal reasoning in safety-critical NLP.

What it makes harder to question

Whether the dataset’s manual annotation process, hierarchical schema, or OSHA source material adequately represent real-world causal complexity for model training.

How the spin works

It combines domain gravitas (construction accidents + OSHA) with methodological signals (manual annotation, hierarchical schema, error analysis) to elevate the dataset’s legitimacy. The framing makes the technical contribution feel larger than warranted by its scale or deployment readiness, while the tension lies between strong claims about 'reliable Causal Information Extraction' and the documented inability of all evaluated models to achieve precise span-level accuracy.

Who Benefits If This Frame Spreads

  • Research authors

    Increased citations and perceived authority in causal NLP and safety-AI subfields

    Framing the dataset as addressing implicit, distributed causality in high-consequence domains elevates its scholarly weight beyond technical novelty.

The Frame

Rigorous academic contribution to safety-critical AI

Missing Context

  • Dataset size (number of reports/annotations)
  • Annotation guidelines or quality control process
  • Potential biases in OSHA reporting practices affecting causal representation

Spin Types

Every story gets a Spin Verdict: a primary spin type (and secondary when the framing blends), a specific tactic name, and a score for how strongly the narrative is steered. Examples beneath each type are tactics, not separate categories.

The Cushion

— Softens negative news

Reframes setbacks, layoffs, delays, losses, or criticism as necessary transitions, efficiency moves, temporary headwinds, or strategic resets — making the downside feel smaller, more acceptable, or less alarming.

Tactics: job-loss softening · restructuring framing · efficiency framing · strategic reset · temporary headwinds

The Shield

— Deflects blame

Shifts responsibility away from the actor — toward regulators, market forces, competitors, bad actors, legacy systems, or abstract risks — while positioning the subject as reactive, responsible, or protective.

Tactics: regulatory blame shift · macroeconomic headwinds · safety framing · bad-actor framing · market-pressure framing

The Hype

— Amplifies future upside

Emphasizes breakthrough potential, massive growth, democratization, transformation, or category disruption while downplaying uncertainty, cost, adoption risk, or timeline friction.

Tactics: innovation framing · democratization · breakthrough framing · category creation · moonshot framing

The Halo

— Associates with virtue primary

Wraps the story in public-good language — responsibility, safety, inclusion, access, sustainability, national interest, or mission — so the subject appears morally aligned and criticism feels harder to make.

Tactics: altruistic reframing · public good · responsible AI framing · inclusion framing · mission-first framing

The Fog

— Obscures details

Uses jargon, passive voice, vague claims, complex phrasing, or missing specifics to make it harder to identify who decided what, what changed, what failed, or what trade-offs were made.

Tactics: strategic ambiguity · jargon saturation · passive voice distancing · accountability blur · undefined metrics

The Stampede

— Creates inevitability

Frames a trend, product, market shift, or decision as already happening, unavoidable, or something everyone must respond to now — creating urgency, FOMO, and pressure to accept the narrative.

Tactics: arms-race framing · inevitability framing · FOMO framing · adoption momentum · future-is-here framing

Spin Score measures how strongly the framing steers the narrative (0–100%). Higher scores mean more deliberate spin tactics — loaded language, selective emphasis, or omitted context. Many stories blend two types (e.g. Halo + Hype).

SpinGraph

How this belief gets built

Claim → Frame → Beneficiary → Gap → AI Risk

The paper frames its contribution as inherently valuable because it tackles causal reasoning in construction safety — a domain where mistakes cost lives — making the dataset feel more urgent and authoritative than generic NLP benchmarks.

  1. Claim

    We introduce ConstructCIE

    We introduce ConstructCIE, a manually annotated dataset for Causal Information Extraction from OSHA construction accident reports.

  2. Frame

    Progress framed as virtuous

    Rigorous academic contribution to safety-critical AI

  3. Beneficiary

    Increased citations and perceived authority in causal NLP and safety-AI

    Research authors — Increased citations and perceived authority in causal NLP and safety-AI subfields

  4. Gap

    Dataset size (number of reports/annotations)

  5. AI Risk

    AI may repeat the headline as fact

    New dataset ConstructCIE helps extract causal factors from construction accident reports, but current LLMs struggle with precise evidence-span identification.

Claim Ledger

01 Primary Product Claim Present in Source risk:Low

We introduce ConstructCIE, a manually annotated dataset for Causal Information Extraction from OSHA construction accident reports.

evidence: Direct self-reporting of dataset creation and source

"We introduce ConstructCIE, a manually annotated dataset for Causal Information Extraction from OSHA construction accident reports."

Evidence Gaps

  • Link to dataset repository
  • Documentation of annotation protocol
  • Inter-annotator agreement score

Fact Check Signals

No direct fact-check match found

0 of 1 claim matched · confidence: low · checked August 10, 2026

01 No direct match

We introduce ConstructCIE, a manually annotated dataset for Causal Information Extraction from OSHA construction accident reports.

Fact Check Signals

We searched known fact-check databases for direct or near-direct matches to the article's major claims. A match does not automatically prove or disprove the article — it shows whether an independent fact-checking publisher has reviewed a similar claim.

  • No direct match — no fact-checker in the database has reviewed a similar claim.
  • Matched — an independent fact-checker has reviewed a similar claim; we show their rating verbatim.
  • Conflicting coverage — fact-checkers disagree on a similar claim.

This is evidence discovery, not an automated truth score. Ratings and wording come directly from the publishing fact-checker.

Language Heatmap

Loaded terms that carry the frame beyond the facts.

ConstructCIE: A Dataset for Extracting Causal Information from Construction Accident Narratives

rich causal information Loaded framing

Carries emotional weight beyond the underlying fact.

reliable Causal Information Extraction Loaded framing

Carries emotional weight beyond the underlying fact.

stronger domain grounding Loaded framing

Carries emotional weight beyond the underlying fact.

Frame Strength

Frame Strength

Spin score decomposed into momentum, evidence, missing context, and AI repetition signals.

Spin Score 35%
Evidence Strength 75%
Narrative Risk 25%
AI Repetition Risk 75%
Missing Context Risk 80%
Virtue / Public Good 60%

Frame Strength Signals

Frame Strength decomposes the overall spin into individual signals. Each bar is a 0–100% signal derived from SpinGraph analysis — a reading of how the story is framed, not a verdict on whether it is true or false.

Reading the ranges

Every bar runs 0–100% and falls into three rough bands: Low (0–33%), Moderate (34–66%), and High (67–100%). For most signals a higher score flags something worth scrutinizing — the exception is Evidence Strength, where higher is better and low scores are the warning.

Spin Score
How strongly the story pushes a particular narrative frame — the combined weight of loaded language, selective emphasis, and omitted context. 0% reads as neutral reporting; higher means more deliberate spin.
  • 0–33% Low — Largely neutral reporting; little detectable framing.
  • 34–66% Moderate — Noticeable slant — the story leans a particular way.
  • 67–100% High — Heavily framed; the angle drives the piece.
Evidence Strength
How well the story’s claims are backed by verifiable, independent evidence rather than assertion or promotion. Higher is stronger. Low scores flag claims that rest on the source’s own word.
  • 0–33% Weak — Claims rest mostly on assertion or a single interested source.
  • 34–66% Mixed — Some verifiable backing, but key claims are thinly sourced.
  • 67–100% Strong — Well supported by independent, checkable evidence.
Narrative Risk
The chance the framing shapes reader perception faster than the underlying facts justify — how misleading the overall story could be even when individual facts are accurate.
  • 0–33% Low — Framing stays close to what the facts support.
  • 34–66% Moderate — Framing outruns the facts in places — read with care.
  • 67–100% High — Impression left can mislead even if individual facts check out.
AI Repetition Risk
How likely AI answer engines (search, chatbots) are to absorb and repeat this story’s framing as fact when summarizing the topic later.
  • 0–33% Low — Framing is unlikely to propagate through AI summaries.
  • 34–66% Moderate — Some risk the slant gets echoed as fact.
  • 67–100% High — Framing is sticky and likely to be repeated as fact.
Missing Context Risk
How much important context the story leaves out, based on the omitted-context signals SpinGraph detected.
  • 0–33% Low — Little material context appears to be omitted.
  • 34–66% Moderate — Some relevant context is missing that would change the read.
  • 67–100% High — Key context is left out, skewing the takeaway.
Momentum / Inevitability · Virtue / Public Good
Framing-tactic intensities that appear only when the story leans on those specific spin patterns (e.g. “the future is already here” or “this is for the public good”).
  • 0–33% Low — The tactic is barely present.
  • 34–66% Moderate — The tactic shapes part of the framing.
  • 67–100% High — The tactic is a dominant part of the pitch.

Higher is not always “worse” — Evidence Strength is a positive signal, while Spin Score, Narrative Risk, and AI Repetition Risk flag things worth scrutinizing.

Reader Risk

What this story makes easy to believe — and what it makes hard to question.

Evidence Strength

Medium

Claims about dataset structure and model performance are internally consistent and methodologically described, but no external validation or replication data is provided.

Verification Status

Claim Present in Source

Narrative Risk

Low

The paper makes modest, empirically bounded claims about model limitations and dataset design; no commercial promises, policy assertions, or safety guarantees are made.

AI Repetition Risk

Moderate

Source Role & Intent

arXiv Computation and Language · Analyst

Intent: Academic Distribution Primary: Announcement Independence: High Spin Weight: Low Trust Weight: High

Counter-Frames

Brand Frame

Rigorous academic contribution to safety-critical AI

Media / Reader Counter-Frame

May be framed as 'AI still can’t parse real-world safety reports' — overemphasizing failure while underrepresenting progress on coarse-grained classification.

Regulatory Counter-Frame

Could prompt scrutiny on whether OSHA’s own reporting formats enable or hinder machine-readable causal analysis — a systemic issue not addressed in the paper.

AI Summary Frame

May conflate 'span-boundary errors' with general hallucination, ignoring the paper’s distinction between hierarchical schema adherence and token-level precision.

Missing Voices

OSHA safety inspectorsconstruction site supervisorsworkers who file incident reports

Questions Not Answered

  • How many annotators were used and what was inter-annotator agreement?
  • What specific OSHA report years or geographies are covered?
  • Were any model failures validated against ground-truth expert review of extracted spans?

Recall Trigger Score

Which stories are likely to become AI memory — separate from Spin Score.

37

Trigger score 30

Not tracked

Triggered by: Major AI entity · Research citation

Not tracked — low-authority source, weak claim, or no durable entity.

AI Recall

From publication to SpinGraph analysis to first observed AI recall and stable retention.

What AI Will Probably Repeat

"New dataset ConstructCIE helps extract causal factors from construction accident reports, but current LLMs struggle with precise evidence-span identification."

Concern: AI may drop the nuance that 'strong accident-type prediction' coexists with 'limited precise span-level extraction', flattening the finding into 'LLMs fail at construction safety tasks'.

  1. Published

    Aug 10, 2026

  2. Ingested

    Aug 10, 2026

  3. SpinGraph Created

    Aug 10, 2026

  4. First Observed AI Recall

    Pending

    Monitoring scheduled

  5. Stable Recall

    —

    Awaiting retention signal

Recall Check Log

No checks yet — recall tracking is opt-in per story.

Sign in to check AI recall

─── GEOGrow AI Recall Layer ───

AI Recall Tracking

Monitoring scheduled. No LLM recall detected yet.

This story has not yet appeared in tested AI answers. Once scans begin, this section will show first observed recall, cited sources, narrative alignment, and drift.

node_id=sts_constructcie_a_dataset_for_extracting_causal_inf

Ask AI about this story

Opens with the SpinGraph .md URL and structured context — one click, prompt included.

ChatGPT Claude Perplexity Gemini Grok

More from arXiv Computation and Language

View all →

Markdown (.md) · JSON-LD schema (.json) · Machine-readable for AI & GEO