Back to Dashboard
Methodology
Understand our data gathering, answer verification, explanation generation, and test engine logic.
1. Data Aggregation
We aggregate questions from 11 community git repositories. Duplicate questions are filtered out using Jaccard similarity.
2. Multi-Role Verification
Answers are cross-referenced with documentation. Mismatches are resolved by community moderators or verified assessment keys.
3. Explanation Curation
We write clear conceptual explanations for each answer. If pseudocode is present, we provide step-by-step variable tracing.
Our Data Processing Pipeline
We process our open source repository files using a strict static processing build step:
- Ingestion: JSON files are parsed during server-side build steps. Questions are cleaned of markdown noise.
- Formatting: Pseudocode logic blocks are detected dynamically and rendered in isolated monospaced code blocks with syntax helpers and line numbers for readability.
- Keyword Indexing: The
key_topics.jsonfile binds related terms (such as EC2, IAM for Cloud) to categories to support quick filtering.
Quiz Engine Scoring Logic
Our practice quiz simulator matches the official Accenture testing conditions:
- Timer: Assessments are timed. The duration changes proportionally with the number of questions.
- Instant Feedback:While the final assessment doesn't show mid-test correctness, our mode immediately checks your selected answer and renders the explanation.
- Scoring: Scores are calculated as:
Correct Answers / Total Questions * 100. Only assessments score ≥ 60% are marked as passed.