Skip to main content
Loading navigation
Back to challenges

Data Center Environmental Impact & Public Hearing Sentiment Monitor

Public opposition to data center development focuses heavily on acoustic noise pollution and local water supply depletion. Build an automated NLP transcript analyzer using Hugging Face Transformers to classify public hearing transcripts into discrete opposition categories (Noise, Water, Grid, Taxes). Achieve 88% macro F1 score across opposition risk classes.

Business OperationsHosted by Vera
Sign in free to start

What you are building

Construct an NLP sentiment and topic classification pipeline using Hugging Face Transformers to index community backlash against data center developments.

How work is evaluated

Evaluation

Evaluates multi-label topic classification accuracy on public testimony transcripts.

Datasets

Public datasets, plus the private datasets you have access to.

How submissions are scored

Dimensions
2 scoring checks
Binary
2 pass or fail dimensions
Ordinal
0 scaled dimensions
Dimension 1

primary_category_match

Verifies primary category classification.

Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Dimension 2

f1_score

Macro F1 score across multi-label opposition categories. • target: 0.88 • range: 0-1

Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Learning goals

  • Fine-tune a Hugging Face BERT/RoBERTa model for multi-label text classification

  • Parse public planning commission transcript text into speaker chunks

  • Classify opposition topics into Noise, Water Usage, Grid Capacity, and Tax Incentives

  • Generate municipal backlash intensity scores for developer risk assessment

Resources

Dataset notes

Dataset of public hearing testimonies, community petition texts, and city council meeting minutes regarding data center approvals.

Challenge setup

Evaluates multi-label topic classification accuracy on public testimony transcripts.

Challenge input

JSON object with transcript_chunk

Transformers

Standard open-source transformer architecture framework.

Evaluated output

JSON object with primary_category, flagged_categories (array of strings)

Evaluation checks
  • Verifies primary category classification.
  • Macro F1 score across multi-label opposition categories. • target: 0.88 • range: 0-1
Configured evaluation evidence
  • Benchmark: NLPTextClassificationBench
  • F1 Score target: 0.88
  • 1 public reference case
Configured runtime
  • Python execution harness
  • Python sandbox (unavailable on Versalist)

Shows the first task, up to three tools, and up to three evidence items.

View technical recipe

Configured tools

Action Space
  • Transformers · Required
  • Hugging Face · Optional
  • LiveKit · Optional

Evaluation contract

  • primary_category_match · Weight 1
  • f1_score · Weight 1

Metrics

  • f1_score · Target 0.88

Benchmarks

  • NLPTextClassificationBench
  • general

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON object with transcript_chunk

    Output: JSON object with primary_category, flagged_categories (array of strings)

    Execution harness: Python

Evaluation tools

  • Transformers · NLP Framework

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Data Center Environmental Impact & Public Hearing Sentiment Monitor