Tokenization basics in Data Analysis Python - Step-by-Step Execution

Concept Flow - Tokenization basics

Start with text string

↓

Split text into tokens

↓

Clean tokens (optional)

↓

Output list of tokens

Tokenization breaks a text string into smaller pieces called tokens, usually words or symbols, to analyze text step-by-step.

Execution Sample

Data Analysis Python

text = "Hello world!"
tokens = text.split()
print(tokens)

This code splits a simple sentence into words (tokens) using space as separator.

Execution Table

Step	Action	Input	Output
1	Start with text string	"Hello world!"	"Hello world!"
2	Split text by spaces	"Hello world!"	["Hello", "world!"]
3	Print tokens	["Hello", "world!"]	["Hello", "world!"]
4	End	-	-

💡 All words separated by spaces become tokens; process ends after printing tokens.

Variable Tracker

Variable	Start	After split	Final
text	"Hello world!"	"Hello world!"	"Hello world!"
tokens	undefined	["Hello", "world!"]	["Hello", "world!"]

Key Moments - 2 Insights

Why does the token 'world!' include the exclamation mark?

Because split() only breaks text by spaces and does not remove punctuation, so 'world!' stays as one token as shown in execution_table step 2.

What if the text has multiple spaces between words?

split() treats any number of spaces as separators, so tokens will still be words without spaces, as split() removes extra spaces automatically.

Visual Quiz - 3 Questions

Look at the execution table at step 2, what is the output of splitting the text?

A ["Hello world!"]

B ["Hello", "world!"]

C ["Hello", "world"]

D ["Hello!", "world"]

💡 Hint

Check the Output column in execution_table row with Step 2.

According to variable_tracker, what is the value of 'tokens' before splitting?

A "Hello world!"

B []

C undefined

D ["Hello", "world!"]

💡 Hint

Look at the 'tokens' row under 'Start' column in variable_tracker.

If the text was "Hi there", how would split() treat the spaces?

A It would create tokens ["Hi", "there"]

B It would create tokens ["Hi", "", "there"]

C It would create one token ["Hi there"]

D It would remove all spaces and create ["Hithere"]

💡 Hint

Recall that split() removes extra spaces and splits on any whitespace.

Concept Snapshot

Tokenization basics:
- Tokenization splits text into tokens (words/symbols).
- Common method: split() breaks text by spaces.
- Tokens may include punctuation unless cleaned.
- Useful for text analysis and processing.

Full Transcript

Tokenization is the process of breaking a text string into smaller parts called tokens, usually words. We start with a text string, then split it by spaces to get tokens. For example, splitting "Hello world!" by spaces gives two tokens: "Hello" and "world!". Note that punctuation like the exclamation mark stays attached to the word because split() only separates by spaces. Variables like 'text' hold the original string, and 'tokens' hold the list of words after splitting. This process is important for analyzing text step-by-step in data science.

Practice

1. What is the main purpose of tokenization in text analysis?

easy

A. To split text into smaller parts called tokens

B. To translate text into another language

C. To remove all punctuation from text

D. To count the number of characters in text

Solution

1. Step 1: Understand tokenization concept

   Tokenization means breaking text into smaller pieces called tokens, like words or sentences.

2. Step 2: Compare options with definition

   Only To split text into smaller parts called tokens correctly describes tokenization as splitting text into tokens.

3. Final Answer:

   To split text into smaller parts called tokens -> Option A

4. Quick Check:

   Tokenization = splitting text [OK]

Hint: Tokenization means splitting text into pieces [OK]

Common Mistakes:

• Confusing tokenization with translation
• Thinking tokenization removes punctuation only
• Mixing tokenization with counting characters

2. Which of the following is the correct way to import the word_tokenize function from nltk in Python?

easy

A. from nltk import word_tokenize.tokenize

B. import nltk.word_tokenize

C. from nltk.tokenize import word_tokenize

D. import word_tokenize from nltk.tokenize

Solution

1. Step 1: Recall correct Python import syntax

   The correct syntax to import a function from a module is 'from module import function'.

2. Step 2: Match syntax with nltk tokenization import

   from nltk.tokenize import word_tokenize matches this syntax exactly for word_tokenize from nltk.tokenize.

3. Final Answer:

   from nltk.tokenize import word_tokenize -> Option C

4. Quick Check:

   Correct import syntax = from nltk.tokenize import word_tokenize [OK]

Hint: Use 'from module import function' to import specific functions [OK]

Common Mistakes:

• Using dot notation incorrectly in import
• Trying to import with 'import function from module'
• Confusing module and function names

3. What will be the output of the following code?

from nltk.tokenize import word_tokenize
text = "Hello, world! Let's tokenize this sentence."
tokens = word_tokenize(text)
print(tokens)

medium

A. ['Hello', ',', 'world', '!', 'Let', "'s", 'tokenize', 'this', 'sentence', '.']

B. ['Hello world Lets tokenize this sentence']

C. ['Hello', 'world', 'Lets', 'tokenize', 'this', 'sentence']

D. ['Hello,', 'world!', "Let's", 'tokenize', 'this', 'sentence.']

Solution

1. Step 1: Understand nltk.word_tokenize behavior

   word_tokenize splits text into words and punctuation as separate tokens, including contractions.

2. Step 2: Analyze the given text tokenization

   It splits "Hello," into 'Hello' and ',', "world!" into 'world' and '!', and "Let's" into 'Let' and "'s" separately.

3. Final Answer:

   ['Hello', ',', 'world', '!', 'Let', "'s", 'tokenize', 'this', 'sentence', '.'] -> Option A

4. Quick Check:

   word_tokenize splits punctuation and contractions separately [OK]

Hint: word_tokenize splits punctuation and contractions separately [OK]

Common Mistakes:

• Assuming punctuation stays attached to words
• Ignoring splitting of contractions
• Expecting a single string token

4. Identify the error in this tokenization code snippet:

import nltk
text = "Data Science is fun!"
tokens = nltk.word_tokenize(text)
print(tokens)

medium

A. word_tokenize is not a function in nltk

B. Missing import of word_tokenize from nltk.tokenize

C. text variable is not defined

D. print statement syntax error

Solution

1. Step 1: Check how word_tokenize is accessed

   word_tokenize is inside nltk.tokenize module, not directly under nltk.

2. Step 2: Identify import mistake

   Importing only nltk does not allow calling nltk.word_tokenize; must import from nltk.tokenize.

3. Final Answer:

   Missing import of word_tokenize from nltk.tokenize -> Option B

4. Quick Check:

   word_tokenize needs correct import [OK]

Hint: Import word_tokenize from nltk.tokenize, not just nltk [OK]

Common Mistakes:

• Calling nltk.word_tokenize without proper import
• Assuming all functions are directly under nltk
• Ignoring module structure

5. You have a list of sentences and want to tokenize each sentence into words, then count total tokens. Which code correctly does this using nltk's word_tokenize?

hard

A. from nltk.tokenize import word_tokenize
sentences = ["I love data.", "Tokenization is key."]
tokens = word_tokenize(sentences)
print(len(tokens))

B. import nltk
sentences = ["I love data.", "Tokenization is key."]
tokens = nltk.word_tokenize(sentences)
print(len(tokens))

C. from nltk.tokenize import word_tokenize
sentences = "I love data. Tokenization is key."
tokens = word_tokenize(sentences)
print(tokens)

D. from nltk.tokenize import word_tokenize
sentences = ["I love data.", "Tokenization is key."]
tokens = [word_tokenize(s) for s in sentences]
total = sum(len(t) for t in tokens)
print(total)

Solution

1. Step 1: Understand input and function usage

   word_tokenize expects a string, so to tokenize multiple sentences, tokenize each string separately.

2. Step 2: Analyze each option's approach

   from nltk.tokenize import word_tokenize sentences = ["I love data.", "Tokenization is key."] tokens = [word_tokenize(s) for s in sentences] total = sum(len(t) for t in tokens) print(total) tokenizes each sentence in a list comprehension, then sums lengths correctly. Some options wrongly pass list to word_tokenize. from nltk.tokenize import word_tokenize sentences = "I love data. Tokenization is key." tokens = word_tokenize(sentences) print(tokens) tokenizes a single string, not list.

3. Final Answer:

   The correct code tokenizes each sentence and counts total tokens -> Option D

4. Quick Check:

   Tokenize each sentence separately, then count tokens [OK]

Hint: Tokenize each sentence string separately, then count tokens [OK]

Common Mistakes:

• Passing list directly to word_tokenize
• Not summing token counts correctly
• Tokenizing entire text as one string when list given