Class 7 · CBSE AI · Strand C — NLP, Vision, and LLMs Deep-Dive
How AI understands context — why 'bank' means two things
The same word can mean different things depending on its neighbours. How AI handles ambiguity. For Class 7.
What this concept actually says
- The same word can have completely different meanings depending on surrounding words — this is called ambiguity
- Modern NLP models use attention mechanisms to weigh which surrounding words matter most for meaning
- Context windows define how much surrounding text a model can 'see' at once
An analogy your child will recognise
Chai at different times of day
If someone says 'Make it strong', you know they mean strong chai in the morning and maybe a strong password in a tech conversation — the surrounding situation changes what 'strong' means completely. That surrounding situation is context.
Cricket commentary
A commentator says 'He's sweeping beautifully!' — in a cricket match that's a batting shot, but in a cleaning store advertisement it's something very different. The stadium, the crowd noise, the previous sentence — all of that is the context that makes meaning clear.
Common misconceptions to watch for
- A larger context window always means better understanding — the model can still 'lose focus' on distant tokens even within its window.
- Context only means the sentence a word is in — in reality, prior conversation, the user's profile, and even the platform the AI is deployed on all constitute context.
Key facts in one breath
- The technical term for this phenomenon is 'lexical ambiguity' — one word form, multiple possible meanings.
- The Transformer architecture (2017) introduced the 'attention mechanism' specifically to solve the context problem by letting each word look at all other words in the sentence.
- GPT-4 has a context window of 128,000 tokens — roughly 100,000 words — allowing it to consider a very large surrounding text.
- Languages like Sanskrit and Tamil have highly context-dependent word meanings, making NLP for these languages especially challenging.
How Dhee Learning teaches this — the 3-stage question loop
Every Dhee Learning session for this concept follows three stages. We share the questions Dhee actually asks, so you can hear what a session sounds like.
Stage 1 — Surface
Read these two sentences: 'She sat on the bank of the river' and 'He withdrew money from the bank.' How does your brain know 'bank' means something different each time — and what clues are you using?
Rote answer
"Context means the words around a word."
Understood
"My brain automatically looks at 'river' and 'withdrew money' nearby and uses those as signals — the rest of the sentence changes what 'bank' means. I don't even decide consciously; it just happens."
Stage 2 — Reasoning
If a machine assigns a fixed embedding to 'bank' regardless of context, what kinds of mistakes would it make — and why is that a serious problem for something like a legal document analyser?
Follow-up Dhee may use: What if you gave every meaning of 'bank' its own separate entry in the dictionary — would that solve the problem? What new problem might it create?
Stage 3 — Application
You're building a chatbot for a cricket academy. A user asks: 'How do I improve my drive?' In cricket, 'drive' is a batting shot. But how might a naive AI get confused, and what context signals would help it get it right?
Misconception Dhee watches for: Assuming that because the chatbot is 'about cricket' it automatically knows all cricket terms — the model still needs explicit context signals to resolve ambiguity.
Related concepts
Want your child to actually understand this?
Dhee turns this concept into a 15-minute spoken session — asking, listening, and probing — so your child builds the idea themselves.
Frequently asked questions
What is context — why 'bank' means different things — explained for kids? +
The same word can mean different things depending on its neighbours. How AI handles ambiguity. For Class 7.
What's the most common mistake children make about this concept? +
A larger context window always means better understanding — the model can still 'lose focus' on distant tokens even within its window.
How does Dhee Learning teach this in a Class 7 session? +
Dhee opens with a question — for example: "Read these two sentences: 'She sat on the bank of the river' and 'He withdrew money from the bank.' How does your brain know 'bank' means something different each time — and what clues are you using?" — listens to your child's answer, then probes the reasoning behind it. The session ends when the child can apply the idea to a brand-new situation, not just recall it.