date: 2025-08-07
topics: [frontend, interview-prep, nlp, semantic-search]
day: 4
Day 4 – Interview Prep + Semantic Search Reading
✅ What I Worked On
- Revised HTML, CSS, and JavaScript interview questions from the GreatFrontEnd blog
- Explored how platforms detect similar questions using semantic search
- Studied the concept of semantic similarity using sentence embeddings
📚 What I Learned
🔹Interview Revision
- JavaScript hoisting, scoping rules, and function types
var, let, and const — scope and hoisting behavior- HTML semantic tags and how they impact SEO/accessibility
- CSS specificity, inheritance, and cascading rules
🔹 NLP & Semantic Search (Reading Only)
- Platforms like Stack Overflow use semantic similarity to detect duplicate questions
all-MiniLM-L6-v2 can embed text into high-dimensional vectors- Compare questions using cosine similarity
- Threshold-based matching helps surface similar existing content before posting
No implementation today — just reading and understanding the concepts.
❌ Blockers
- Still gaining clarity on:
- When to consider two questions "similar enough"
- Scaling similarity search in a real project
🧠 Reflection
Revisiting frontend questions sharpened my interview focus.
The semantic search concept looks powerful for DevFlow’s future roadmap, especially to reduce question duplication.
🔗 References