Description
Task 4: “Recommender” (8%). Recommendations are everywhere: Amazon recommends books you
should buy, Netflix tells you which TV series you should watch, and Google recommends news articles to
read. Time to write your own C++ program that can recommend items!
For this assignment, you have to work on movie recommendations (though the program really works with
any content). You have to be able to:
1. Read a set of movie descriptions into an index;
2. Ask a user about a movie s/he likes (which must be one in the index);
3. Recommend a list of top-n other movies that s/he would also like.
Dataset. Your program has to work with the CMU movie dataset,1 which you can download from
http://www.cs.cmu.edu/∼ark/personas/data/MovieSummaries.tar.gz. This dataset contains 42,306 movies
in the file movie.metadata.tsv, including a summary for each movie that you can find in the file
plot summaries.txt. The important parts in these files for the purpose of this assignment are (details are
in the README.txt file included with the dataset):
File movie.metadata.tsv:
Wikipedia movie ID Freebase movie ID Movie name Release Date …
975900 /m/03vyhn Ghosts of Mars 2001-08-24 …
…
File plot summaries.txt:
Wikipedia movie ID Plot description
975900 Set in the second half of the 22nd century…
Indexing. Use your classes from the first three assignments to build an index, where each movie m is
represented through a vector ~m that you create from the movie’s plot description.
Processing the query. Your system takes as input the name of a movie. If the movie exists in the index,
you compute the similarity between the input movie’s vector and the other movies in the index, using the
same similarity function as before.
Recommendation. Output the top-n best (default 5) movies as recommendation (include movie name,
release date, ID#, and plot description in the output).
Congratulations, you now have your very own recommender system. The method used here is called
content-based recommendation, which is one of many recommendation methods in use today.2 Another
approach is collaborative filtering, which you might have heard about in the context of the 2009 competition
organized by Netflix for the best movie rating prediction,3 where the winner was awarded a prize of
US $1,000,000.
1CMU Movie Summary Corpus, http://www.cs.cmu.edu/∼ark/personas/
2
If you want to learn more about recommendation engines, you can start with the Wikipedia page https://en.wikipedia.
org/wiki/Recommender system and for a lot more details, download Recommender Systems: The Textbook through the
Concordia Library: http://mercury.concordia.ca/record=b3281624.
3The Netflix Prize, see https://en.wikipedia.org/wiki/Netflix Prize
COMP 345 Fall 2017 Assignment #4
Coding guidelines. Develop your program according to the following specification:
a) For all classes, make sure you properly separate your system into header (.h) and implementation (.cpp)
files. Put each class into its own translation unit. You are free in the choice of an IDE, but your code must
be standard, cross-platform C++ code.
b) Document all your classes and functions with Doxygen.
c) For your classes, follow object-oriented design principles as discussed in the course; in particular make
data members private unless you have a good reason not to; use friend functions where appropriate to
access private members; access private members in derived classes through protected functions, and make
proper use of inheritance (e.g., use virtual functions for polymorphism and do not override non-virtual
functions in publicly derived classes). New: for all classes, make use of exception handling for any error
cases.
d) Write four separate main programs, using the same classes (see e) below): a new recommender.cpp for
Task 4, as well as updated versions of summarizer.cpp (Task 3), googler.cpp (Task 2) and indexing.cpp
(Task 1). You will have to demo all four main programs.
e) Add these two new classes to your code from the previous assignments:
Class movie is a new subclass of index item that holds the information for a movie: ID, name, and
release date, together with accessor methods. The content field holds a movie’s plot description.
Class index exception is a new exception class, which is a subclass of the standard class std::exception.
Override the virtual function what() to provide an explanation of the exception. Change any existing
code in your classes to use the new exception class in case of errors. Make sure you process the
exceptions (try…catch) so that your main program does not terminate. In particular, for Task 4,
asking for recommendations based on a movie name that is not in the index will result in an exception.
Note that you will have to find a way to combine the information about a movie (name and plot) through
the movie ID from the two input files.
You code must make use of polymorphism, where appropriate. For all these classes, overload the inserter
(operator<<) to provide meaningful debug output. You can add additional classes if you like, but these must
not duplicate the functionality of the classes above. As for the previous assignments, you are responsible for
coming up with an object-oriented design that makes good use of these classes, so that they collaboratively
solve the stated tasks (e.g., using the mentioned CRC method).
2