International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 10 Issue: 05 | May 2023
p-ISSN: 2395-0072
www.irjet.net
Music Recommendation System Akrit Sood1, Uday Sharma2, Harashleen Kour3 1 Department of Computer Science and Engineering, Chandigarh University, Mohali, Punjab, India
2 Department of Computer Science and Engineering, Chandigarh University, Mohali, Punjab, India
Department of Computer Science and Engineering, Chandigarh University, Mohali, Punjab, India ---------------------------------------------------------------------***--------------------------------------------------------------------1. Collaborative filtering: The notion behind filtering is to Abstract - This research paper presents a study on 3
identify people with similar musical interests like similar songs. The above algorithm generates suggestions according to a user's preferences or playing history, albums, ratings, and interests of other users with similar musical likes.
developing a machine learning-based system to provide suggestions for music, utilizing a dataset from Asia's leading music streaming service. The purpose is the study to build a better music system for suggestions and provides personalized recommendations for listeners based on their previous listening behavior. The proposed approach employs both content-based as well as collaborative filtering approaches to produce suggestions. The content-based approach analyzes the properties associated with music, such as genre, tempo, and melody, to find similar songs. The collaborative filtering approach uses user behavior data to recognize other people that have similar hobbies and music preferences and recommends songs that they have listened to. The paper presents the planning and carrying out of the system for a song suggestion, including the data collection, preprocessing, and feature extraction steps. The system is evaluated using the dataset from the music streaming service and compared to a number of baseline algorithms. The conclusions show if the suggested system exceeds the baseline algorithms in relation to recommendation accuracy and diversity. This paper ends with a discussion of conceivable applications and limitations in terms of the planned music recommendation system, as well as future directions for investigating this field. In general, the research demonstrates the effectiveness of methods of learning from machines for building better suggestions for song systems which can improve the music experience with hearing for users.
2. Content-based filtering: To provide recommendations, this filtering concentrates on the qualities of the song itself, such as genre, tempo, and mood. 3. Hybrid systems: To deliver more reliable recommendations, hybrid systems integrate collaborative or content-based filtering. The algorithms used towards creating suggestions decide their effectiveness for these systems. To assess user information and provide recommendations, machine learning methods neural networks for learning, and trees of decision trees are often used. To increase their correctness over time, these algorithms are trained on huge databases of music and consumer preferences. Unfortunately, the data provided can restrict the use of these systems. For example, if a user has only listened to a couple of songs or has a small playlist, the resulting recommendations may be inaccurate. Aside from their usefulness, music recommendation systems involve ethical considerations such as confidentiality and algorithmic unfairness. Some customers may be worried about the volume of private information gathered for recommendation purposes. Despite these reservations, music recommender systems have had a tremendous impact on both the music world and the people they serve. They allow fans to discover additional music that they might not have discovered else, while also introducing artists to new listeners. As these systems develop, it will be essential to find an equilibrium between their effectiveness and ethical issues to ensure that they deliver real benefits to both audiences and artists
Key Words: Music Recommendation System, Machine Learning, Data Pre-processing, Population model, Collaborative Filtering Model, Content-Based Mode, Clustering
1. INTRODUCTION In recent decades, music recommendation algorithms have grown into essential components for music fans. With the development of services that stream music, the volume of music accessible to customers is huge and overwhelming, which makes it hard to discover new artists and songs that match their interests. Music recommendation systems provide a solution by making individualized suggestions based on the listener's interests, listening history, and other criteria. Music recommendation systems are categorized as follows:
© 2023, IRJET
|
Impact Factor value: 8.226
2. LITERATURE SURVEY H. Ying et al., (2018) [1], propose a novel approach to building recommender systems for music by combining matrix factorization, recurrent neural networks, and attention based models. The proposed method uses a hierarchical attention network that can capture both the hierarchical structure of music metadata and the sequential patterns of user behavior. The experimental results show that the proposed method outperforms several state-of-the-
|
ISO 9001:2008 Certified Journal
|
Page 1566