Music Recommendation System for Million Song Dataset Challenge
نویسنده
چکیده
In this paper a system that took 8th place in Million Song Dataset challenge is described. Given full listening history for 1 million of users and half of listening history for 110000 users participatints should predict the missing half. The system proposed here uses memory-based collaborative filtering approach and user-based similarity. MAP@500 score of 0.15037 was achieved.
منابع مشابه
Deep content-based music recommendation
Automatic music recommendation has become an increasingly relevant problem in recent years, since a lot of music is now sold and consumed digitally. Most recommender systems rely on collaborative filtering. However, this approach suffers from the cold start problem: it fails when no usage data is available, so it is not effective for recommending new and unpopular songs. In this paper, we propo...
متن کاملJuggle: large-scale discovery in music recommendation
Today’s offer of audio content exceeds the human capability of manually searching datasets with hundreds of songs, demanding automated tools capable of handling music recommendation when faced with large-scale collections. In this work, we address the playlist generation and song discovery tasks with large-scale datasets. It is possible to quickly obtain playlists and explore collections with e...
متن کاملA Machine Learning based Music Retrieval and Recommendation System
In this paper, we present a music retrieval and recommendation system using machine learning techniques. We propose a query by humming system for music retrieval that uses deep neural networks for note transcription and a note-based retrieval system for retrieving the correct song from the database. We evaluate our query by humming system using the standard MIREX QBSH dataset. We also propose a...
متن کاملAddressing cold-start problem in music recommendation with HMM-based Collaborative Profile Space analysis
In this paper, Collaborative Profile Space analysis, a novel approach based on HMM, is proposed to tackle the “new item” cold start problem in music recommendation task. By calculating the probability of generating a particular song from a set of trained HMMs, coordinates in the Collaborative Profile Space is defined, and used in the classifiers. With self-collected dataset and evaluation strat...
متن کاملContent-Aware Collaborative Music Recommendation Using Pre-trained Neural Networks
Although content is fundamental to our music listening preferences, the leading performance in music recommendation is achieved by collaborative-filtering-based methods which exploit the similarity patterns in user’s listening history rather than the audio content of songs. Meanwhile, collaborative filtering has the well-known “cold-start” problem, i.e., it is unable to work with new songs that...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
- CoRR
دوره abs/1209.3286 شماره
صفحات -
تاریخ انتشار 2012