Natural Language Processing

Fit characters into a small table

Two characters share an address

Botie's feature table has four available buckets even as new characters arrive. A repeatable address rule fits the storage budget but can merge distinct identities.

BOTIE

The second case maps b and f to the same bucket. Which design tradeoff would matter for Botie's next use?

The inputs we are comparing

Character features

Example 1

aec

Example 2

bfg

Result for example 1

Choose an explanation to test

Keep four buckets when fixed storage matters more than exact identityUse an identity-preserving vocabulary when distinguishing every character matters

An emoji and a letter share the same bucket rule

This check uses the pieces from “Map character codes into four buckets”.

BOTIE

Read each Unicode code point, then assign its numerical value modulo four to a bucket. Return one bucket per character, including the emoji as one code point.

Character features: 🙂A

What result should your machine produce for this case?

Use a number, text, or an array matching the requested result. For a matrix, put each row in brackets. Saving your prediction runs your current construction on these inputs and checks the earlier examples too.

Build a construction in the workspace above first. This case will test those pieces.

Could the bucket stream alone reconstruct which of two colliding characters appeared?

Your notes for later. These stay with this journey.

A bounded feature address may sacrifice identity; this code-point remainder is a small demonstration rather than a general-purpose hash function.

This is a small construction problem. The lesson explains the assumptions and the parts of the full method that this example leaves out.

Work through the Python challenges →

Continue the coursePatching Without a Vocabulary →

← Finite-State Morphology