Java-Threaded-Word-Counter

• This repository houses the code used to compare a single-threaded word counter against a multithreaded word counter that I did as an assignment for my Principles of Operating Systems course.

Introduction

• Work on creating word counters through two different implementations and compare their performance
  • Single Threaded Word Counter
  • Multithreaded Word Counter
• The dataset used for this comparison is a sample set of 14 different files containing csv and txt files, provided in the dataset directory
• Hash Maps are used in order to store the key value pairs of word, count
  • For the multithreaded approach I use the Collections.synchronizedMap method from the Collections package

1. Single-Threaded Word Counter

• Implement a single-threaded Java program that will find the most frequent word for each file in the 14-file dataset
• Important Notes:
  • Iterate through the 14 files a total of 3 times
    • 1^st iteration: count the most frequent word with exactly 6 characters
    • 2^nd iteration: count the most frequent word with exactly 7 characters
    • 3^rd iteration: count the most frequent word with exactly 8 characters
  • Print to the console the file names and the most frequent word for each file
  • Make sure to remove case sensitivity and use a reqular expression to identify words
    • Code should ignore whitespace and punctuation
  • Time all iterations and print the total time to the console
• Deliverable:
  • Java source
  • PDF of the screenshot of the word counts and the runtime printed to the console

2. Multithreaded Word Count

• Created a new program that modified the single-threaded program to be multithreaded
• New program must be at least 10% faster than the original single-threaded program
• Implementation includes creating a thread for each file and that thread will be responsible for their own file and completing the word count

Example Output

• The output to the console and the screenshot we must include must look similar to the below image: