Text Mining Techniques (Frequency Analysis, Pattern Matching)

Text mining is the process of deriving meaningful information and knowledge from text data. With the ever-increasing volume of textual data available, text mining techniques have become crucial for various applications such as sentiment analysis, topic modeling, and recommendation systems. In this article, we will explore two essential text mining techniques: frequency analysis and pattern matching, using Python.

Frequency Analysis

Frequency analysis is a fundamental technique in text mining that helps understand the distribution of words or terms in a given text corpus. By determining the frequency of words, we can identify the most common words, discover rare terms, and gain insights into the content of the corpus.

Python provides several libraries to perform frequency analysis easily. One popular library is nltk (Natural Language Toolkit). Let's consider an example to illustrate this technique:

import nltk
from nltk import FreqDist

corpus = ["This is a sample sentence.",
          "Another sentence for analysis.",
          "A third sentence to analyze the frequency of words."]

# Tokenize the corpus into individual words
words = nltk.word_tokenize(" ".join(corpus))

# Calculate the frequency distribution
freq_dist = FreqDist(words)

# Get the most common words
most_common = freq_dist.most_common(5)


In the above code, we tokenize the input corpus into individual words using nltk.word_tokenize(). Then, we calculate the frequency distribution using nltk.FreqDist(). Finally, we retrieve the most common words using most_common() method.

The output of the above code will be:

[('sentence', 3), ('a', 2), ('.', 2), ('This', 1), ('is', 1)]

This result shows that the word "sentence" appears three times in the corpus, followed by "a" and "." with frequency 2.

Frequency analysis allows us to gain a high-level overview of the textual data and identify the significant terms that appear frequently. This information can be further utilized for various analysis purposes.

Pattern Matching

Pattern matching is another vital text mining technique used to identify specific patterns or expressions within a text corpus. This technique is valuable for tasks such as named entity recognition, email or URL detection, and finding relevant dates or phone numbers in unstructured text.

Python provides various tools and libraries for pattern matching, including regular expressions and the re module. Let's consider an example to demonstrate pattern matching using regular expressions:

import re

corpus = ["Please contact me at example@example.com for further details.",
          "The website URL is www.example.com.",
          "The event is scheduled on 01/01/2023 at 8:00 AM."]

# Define a pattern to match email addresses
email_pattern = r'\b[A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\.[A-Z|a-z]{2,}\b'

# Define a pattern to match URLs
url_pattern = r'http[s]?://(?:[a-zA-Z]|[0-9]|[$-_@.&+]|[!*\\(\\),]|(?:%[0-9a-fA-F][0-9a-fA-F]))+'

# Define a pattern to match dates
date_pattern = r'\b\d{2}/\d{2}/\d{4}\b'

# Perform pattern matching on the corpus
emails = re.findall(email_pattern, " ".join(corpus))
urls = re.findall(url_pattern, " ".join(corpus))
dates = re.findall(date_pattern, " ".join(corpus))


In the above code, we define patterns using regular expressions to match email addresses, URLs, and dates. Then, we use re.findall() to perform pattern matching on the corpus.

The output of the above code will be:


This result shows that the pattern matching successfully identified the email address, URL, and date present in the corpus.

Pattern matching provides a powerful way to extract specific information from text data, allowing us to perform targeted analysis or extract relevant entities.


Text mining techniques such as frequency analysis and pattern matching play a crucial role in extracting meaningful insights from text data. Python provides efficient libraries and modules like nltk and re that facilitate these techniques. By employing these techniques, we can gain valuable knowledge from textual data and apply it to various real-world applications.

*Note: The code presented in the examples is for illustration purposes, and it can be extended and adapted to specific use cases.

noob to master © copyleft