To compute the conditional probabilities you need to determine unigram and bigram counts first (you can do this in a single pass through a file if you do things carefully) and store them in a Binary...

To compute the conditional probabilities you need to determine unigram and

bigram counts first (you can do this in a single pass through a file if you do things

carefully) and store them in a Binary Search Tree (BST). After that, you can compute

the conditional probabilities.
Input files
Test files can be found on (http://www.gutenberg.org/ebooks/). For example,

search for “Mark Twain.” Then click on any of his books. Next download the “Plain

Text UTF-8” format.
In addition, you should test your program on other input files as well, for which you

can hand-compute the correct answer.
Output files
Your program must accept the name of an input file as a command line argument.

Let's call the file name of this file fn. Your program must then produce as output the

following set of files:
• Your program must write the unigram counts to a file named fn.uni in which

each unigram is listed on a separate line, and each line contains just the

unigram and its count (an integer), separated by a single space.
• Your program must write the bigram counts to a file named fn.bi in which

each bigram is listed on a separate line, and each line contains just the

bigram and its count (an integer), separated by a single space.
• Your program must write the conditional probabilities to a file named fn.cp,

reported in the form P(WORD(k)|WORD(k-1)) = p, where p is the conditional

probability of WORD(k) given WORD(k-1).
Notes
• You may use any BST implementation found online at your own risk
(provided the source of this code is cited properly). 50 marks bonus would
be given if you implemented the BST yourself (only the functionalities

needed to complete your project).
• Your program should accept file name(s) as command line argument(s) (no

hard-coded file names in your code).
• Your code will be tested using the latest version of the GNU C compiler on a

Unix-based operating system. If you do not have a Unix-based machine you

may want to use Windows Subsystem or a Virtual Box under your own

responsibility!
• Your code must be well commented. When writing your comments, you

should focus on what the code does at a high level; for example, describe the

main steps of an algorithm—not every detail (that what the code is for).
• A ReadMe.txt file (including instructions on how to compile and how to run

your program along with any known problems) must be submitted

Extracted text: Submission details: Make a directory and name it, using your first and last names, applying the camel notation, for example haniGirgis. Please name the main file, where execution starts, as hw2.cpp. Compress your directory by the zip utility, and submit the compressed file using Blackboard. Introduction Suppose you had a document (perhaps a newspaper article, or an unattributed manuscript for a book), and you were interested in knowing who wrote it. One way to try to determine the authorship of the anonymous document is by comparing properties of the anonymous document with properties of known documents, and seeing if there is enough similarity to make a judgment of authorship. Some simple properties one might use to distinguish different authors include: Vocabulary (ie. the set of words an author uses). • Word frequencies (i.e. the frequencies with which an author uses words). Bigram frequencies (i.e. the frequencies of two consecutive words). Bigram probabilities (i.e. the probability that one word follows another word). Terminology A unigram is a sequence of words of length one (i.e. a single word). • A bigram is a sequence of words of length two. • The conditional probability of an event E2 given another event E1, written p(E2|E1), is the probability that E2 will occur given that event E1 has already occurred. We write p(w(k)]w(k-1)) for the conditional probability of a word w in position k, w(k), given the immediately preceding word, w(k-1). You determine the conditiona probabilities by determining unigram counts (the number of times each word appears, written c(w(k)), bigram counts (the number of times each pair of words appears, written c(w(k-1) w(k)), and then dividing each bigram count by the unigram count of the first word in the bigram: p(WORD(k)|WORD(k-1)) = c(WORD(k-1) WORD(k}) / {WORD(k-1)) For example, if the word "time" occurs seven times in a text, and "time of" occurs three times, then the probability of "of occurring after "time" is 3/7. Project description In this project, you will be determining conditional probabilities of bigrams. To do this, you will write a C program, which reads in a file of text and produces three output files, as described below.