A ThesisPresented toThe Faculty of the Department of Computer ScienceSan José State

UniversityIn Partial Fulfillmentof the Requirements for the DegreeMaster of

SciencebyYunkyoung LeeDecember 2015

©2015 Yunkyoung Lee All Rights Reserved

The Designated Project Committee Approves the Project TitledRecommendation System

Using Collaborative FilteringByYunkyoung Lee APPROVED FOR THE DEPARTMENT OF
COMPUTER SCIENCESAN JOSE STATE NIVERSITYDecember 2015Dr. T Y Lin Department of
Computer ScienceDr. H. Chris Tseng Department of Computer ScienceDr. Thomas Austin
Department of Computer Science

ABSTRACT
Recommendation System Using Collaborative Filteringby Yunkyoung Lee
Collaborative filtering is one of the well known and most extensivetechniques in
recommendation system its basic idea is to predict which items auser would be
interested in based on their preferences. Recommendationsystems using collaborative
filtering are able to provide an accurate predictionwhen enough data is provided,
because this technique is based on the user’spreference. User-based collaborative
filtering has been very successful in thepast to predict the customer’s behavior as
the most important part of therecommendation system. However, their widespread use
has revealed some realchallenges, such as data sparsity and data scalability, with
gradually increasingthe number of users and items.To improve the execution time and
accuracy of the prediction problem,this paper proposed item-based collaborative
filtering applying dimensionreduction in a recommendation system. It demonstrates
that the proposedapproach can achieve better performance and execution time for
therecommendation system in terms of existing challenges, according to
evaluationmetrics using Mean Absolute Error (MAE).
ABOUT AUTHOR

Nancy Naina
1.0
PAPERS
1
VIEWS
48
FOLLOWERS
0
Follow
RELATED PAPERS
An Improved Web Page Recommendation System Using Partitioning and Web Usage Mining
Annappa B • 15 Views
DownloadMore Options
Paper Thumbnail
Workshop on recommendation utility evaluation
Xavier Amatriain • 113 Views
DownloadMore Options
Paper Thumbnail
A Content Boosted Hybrid Recommender System
Seval Çapraz • 144 Views
DownloadMore Options
Paper Thumbnail
User Preference Elicitation, Rating Sparsity and Cold Start
Mehdi Elahi • 74 Views
DownloadMore Options
Paper Thumbnail
An Improved Switching Hybrid Recommender System Using Naive Bayes Classifier and
Collaborative Filtering
Kukuh Susanto • 230 Views
DownloadMore Options
Paper Thumbnail
PAPER SURVEY AND EXAMPLE OF COLLABORATIVE FILTERING IMPLEMENTATION IN RECOMMENDER
SYSTEM
Muhammad Hanafi • 44 Views
DownloadMore Options
Paper Thumbnail
A Bray-Curtis strategy for boosting product recommendation
Marcelo Mendoza • 8 Views
DownloadMore Options
Paper Thumbnail
A Hybrid Approach for Improving Prediction Coverage of Collaborative Filtering
K.G. MARGARITIS • 21 Views
DownloadMore Options
Paper Thumbnail
A Survey of Challenges in Collaborative Filtering Recommender System
divya Kannan • 363 Views
DownloadMore Options
Paper Thumbnail
Differential Context Modeling in Collaborative Filtering
Bamshad Mobasher and Robin Burke • 9 Views
DownloadMore Options
Paper Thumbnail
Behavior Based Adaptive Navigation Support
Maria Bielikova • 49 Views
DownloadMore Options
Paper Thumbnail
AN UNDERSTANDING AND APPROACH SOLUTION FOR COLD START PROBLEM ASSOCIATED WITH
RECOMMENDER SYSTEM: A LITERATURE REVIEW HANAFI, 3 NANNA SURYANA, 4 ABD SAMAD BIN
Muhammad Hanafi • 155 Views
DownloadMore Options
Paper Thumbnail
A Review Paper on BFO and PSO Based Movie Recommendation System
J4R - Journal for Research and Bhupender Sharma • 67 Views
DownloadMore Options
Paper Thumbnail
The Use of the Genetic Algorithms in the Recommender Systems
Bushra Alhijawi • 281 Views
DownloadMore Options
Paper Thumbnail
EMERGING TOPICS IN COMPUTING ClubCF: A Clustering-Based Collaborative Filtering
Approach for Big Data Application
Globalirrigation Systems • 9 Views
DownloadMore Options
Paper Thumbnail
COSTOMIZATION OF RECOMMENDATION SYSTEM USING COLLABORATIVE FILTERING ALGORITHM ON
CLOUD USING MAHOUT
eSAT Journals • 22 Views
DownloadMore Options
Paper Thumbnail
The Role of Emotions in Context-aware Recommendation
Bamshad Mobasher • 27 Views
DownloadMore Options
Paper Thumbnail
Comparison of Collaborative Filtering Algorithms with Various Similarity Measures
for Movie Recommendation
Taner Arsan • 21 Views
DownloadMore Options
Paper Thumbnail
Dataset-driven research for improving recommender systems for learning
Riina Vuorikari • 21 Views
DownloadMore Options
Paper Thumbnail
Expert Systems With Applications A collaborative filtering method for music
recommendation using playing coefficients for artists and users
Sohaib Ahmad • 6 Views
DownloadMore Options
Paper Thumbnail
A collaborative filtering method based on artificial immune network
Ahmet Arslan • 81 Views
DownloadMore Options
Paper Thumbnail
A holographic associative memory recommender system
Matthew Rutledge-Taylor • 6 Views
DownloadMore Options
Paper Thumbnail
Fuzzy Logic Based Recommender System
Research and Scientific Innovation Society RSIS International • 77 Views
DownloadMore Options
Paper Thumbnail
Expert Systems With Applications A recommender system based on collaborative
filtering using ontology and dimensionality reduction techniques
Lovepreet Sangwan • 31 Views
DownloadMore Options
Paper Thumbnail
Investigating the Scalability of Algorithms, the Role of Similarity Metric and the
List of Suggested Items Construction Scheme in Recommender Systemsfree from
spelling and grammatical errors
and is written in a professional style of a technical
article and presented at a high standard.
A+
95
C2 – Work Showing Evidence:
1. The introduction explains the topic at an
outstanding level and the problem domain has been
discussed very clearly. An outstanding level of
understanding of the topic material has been
shown.
2. Outstanding research work has been done with
proper explanations using proper sources and
referencing. An outstanding level of understanding
has been shown.
3. The proposed solution has been described perfectly
using necessary diagrams and clear pseudocode. An
outstanding level of understanding of the solution is
displayed.
4. The report is well structured, written well, and free
from spelling and grammatical errors and is written
at a very good quality standard.YouTube
are affording users with access to songs to their paid
members. It is important that these companies need to
maintain their members to keep a subscription to gain
more revenue. The playlist is a special function of these
streaming apps. Many users feel difficult to create a list
from a long list of music. As a result, users tend to play
next song by in a random mode or by recommendation.
Hence, a successful personalized music recommendation
technique has become key to stay their
members from jumping to another service. Techniques of
music recommendation have been developed in the
previous decade.
Collaborative filtering has been widely used in previous
research. In addition to that, Millions of Songs data
challenge has attracted many researchers to work in this
area. Recent years, Artificial neural network (ANN)
model has made advancement in sequence data
classification. The above deep learning model has
advanced in natural language processing. Machine
translation is a successful application. Also, in the field
of computer vision, has a good result in sequence video
analysis ANN model and KNN Regression algorithm is
used to make comparisons of different songs by
similarity, which will help the recommendation system
to give ranking score solely based on combination
factors of music themselves.
II. RELATED WORK
Very few works of literature have been done to apply for
deep learning framework in music recommendation,
especially on audio. Music recommender systems
(MRSs) have experienced a sudden growth in recent
years, with the increase of online streaming services.
Collaborative filtering approaches are used where the
songs are recommended based on several user’s
preferences and user’s listening history. Content based
filtering approaches are used where the songs are
recommended based solely on the content. There are
hybrid filtering approaches which uses both user’s
preferences and content of songs.
III. MODEL AND APPROACHES
A. Approaches
The approach we use focuses on various audio
features and the lyrics scores are taken for each pair of
songs. To make it more specific, consider two different
songs P and Q. P is represented as a set of attributes
(p1,p2,…..pn) and Q is represented as a set of attributes
(q1,q2,…qn). Then the model is trained using these two
vectors, which element is a chunk of music. Using the
formula: y=f (P, Q), where, y means
the similarity score between two songs, and f is a
similarity measurement function. By treating this function
as a model, the RNN model and KNN Regression model
is used to train the predicted model.
B. Model
The proposed model we used is based on Artificial neural
network and KNN Regression algorithm. The motivation
behind using an ANN-based architecture stems from the
fact that the similarity score between songs is based on
various deep audio features of the song along with the
lyrics used in the song. KNN regression is good at
predicting similarities between the pair of the song given
the attributes. Web scrapping is used to download music
files, thumbnails and artist images from various web
International Journal of Engineering Research & Technology (IJERT)
ISSN: 2278-0181 http://www.ijert.org
IJERTV8IS070267
(This work is licensed under a Creative Commons Attribution 4.0 International
License.)
Published by :
www.ijert.org
Vol. 8 Issue 07, July-2019
864
sources. The above data is uploaded to firebase storage
and firestore and made available using a web user
interface. ANN model that has one input layer, six
hidden layers, and one output layer. A normal initializer
is used to initialize the weights. All hidden layer uses the
RELU activation function. The neuron sizes at hidden
layers are 12,22,12,22,12,8 respectively. The output layer
uses the linear activation function. The model uses mean
squared error as the loss function. The optimizer used is
ADAM optimizer. 200 epochs were used to train the
model. KNN Regression uses 30 neighbour points to
predict the similarity scores. The metric used for
evaluation of the model is mean squared error.
C. Dataset
In this paper, we use three datasets which are Million
Song dataset, Musixmatch dataset, and Lastfm dataset.
Million song dataset contains audio features and
metadata of each song. The million song dataset is in
HDF5 format. Features are extracted from this dataset.