Apprendre Working with Text Data in PySpark

Glissez pour afficher le menu

Text columns require their own preprocessing pipeline before they can be used in a model. Raw strings need to be split into tokens, cleaned, and eventually converted into numeric vectors.

The flights dataset does not contain free-text fields, but CANCELLATION_REASON is a categorical code. To demonstrate text processing, you will work with a small synthetic dataset of flight status descriptions alongside the main dataset.

Creating a Text DataFrame


              1234567891011121314151617181920212223
            
import urllib.request
from pyspark.sql import SparkSession

urllib.request.urlretrieve(
    "https://staging-content-media-cdn.codefinity.com/courses/aa80ac56-0d50-49e8-9231-2c2374cd3e9d/flights.csv",
    "flights.csv"
)

spark = SparkSession.builder \
    .appName("TextData") \
    .master("local[*]") \
    .getOrCreate()

# Synthetic flight status descriptions
text_df = spark.createDataFrame([
    (1, "flight delayed due to bad weather conditions"),
    (2, "technical issue caused significant departure delay"),
    (3, "flight cancelled due to severe weather"),
    (4, "late aircraft arrival caused departure delay"),
    (5, "air traffic control delay affected departure time"),
], ["id", "description"])

text_df.show(truncate=False)

Tokenization

Tokenizer splits a string into lowercase tokens by whitespace:


              123456
            
from pyspark.ml.feature import Tokenizer

tokenizer = Tokenizer(inputCol="description", outputCol="tokens")
tokenized_df = tokenizer.transform(text_df)

tokenized_df.select("description", "tokens").show(truncate=False)

Removing Stop Words

Common words like "due", "to", and "a" carry little meaning. StopWordsRemover filters them out:


              123456
            
from pyspark.ml.feature import StopWordsRemover

remover = StopWordsRemover(inputCol="tokens", outputCol="filtered_tokens")
filtered_df = remover.transform(tokenized_df)

filtered_df.select("tokens", "filtered_tokens").show(truncate=False)

Tout était clair ?

Merci pour vos commentaires !

Section 1. Chapitre 5

Demandez à l'IA

Posez n'importe quelle question ou essayez l'une des questions suggérées pour commencer notre discussion

Section 1. Chapitre 5