Time-Based K-nearest Neighbor Collaborative Filtering

2012 ◽  
Author(s):  
Yue Liu ◽  
Zhe Xu ◽  
Binkai Shi ◽  
Bofeng Zhang
Keyword(s):  
Collaborative Filtering ◽  
Nearest Neighbor ◽  
K Nearest Neighbor

Song Recommendations Based on Artists with Cosine Similarity Algorithms and K-Nearest Neighbor

2020 ◽  
Vol 8 (4) ◽  
pp. 367
Author(s):  
Muhammad Arief Budiman ◽  
Gst. Ayu Vida Mastrika Giri
Keyword(s):  
Collaborative Filtering ◽  
Mobile Phones ◽  
Recommendation System ◽  
Nearest Neighbor ◽  
Cosine Similarity ◽  
Music Recommendation ◽  
K Nearest Neighbor ◽  
Filtering Method ◽  
K Nearest Neighbor Algorithm ◽  
Music Recommendation System

The development of the music industry is currently growing rapidly, millions of music works continue to be issued by various music artists. As for the technologies also follows these developments, examples are mobile phones applications that have music subscription services, namely Spotify, Joox, GrooveShark, and others. Application-based services are increasingly in demand by users for streaming music, free or paid. In this paper, a music recommendation system is proposed, which the system itself can recommend songs based on the similarity of the artist that the user likes or has heard. This research uses Collaborative Filtering method with Cosine Similarity and K-Nearest Neighbor algorithm. From this research, a system that can recommend songs based on artists who are related to one another is generated.

Collaborative filtering recommendation algorithm based on class correlation distance

2019 ◽  
Vol 13 ◽  
Author(s):  
Hanfei Zhang ◽  
Yumei Jian ◽  
Ping Zhou
Keyword(s):  
Collaborative Filtering ◽  
Nearest Neighbor ◽  
Experimental Result ◽  
K Nearest Neighbor ◽  
Recommendation Algorithm ◽  
K Value ◽  
Correlation Distance ◽  
Training Samples ◽  
The Difference ◽  
Better Than

: A class correlation distance collaborative filtering recommendation algorithm is proposed to solve the problems of category judgment and distance metric in the traditional collaborative filtering recommendation algorithm, which is using the advantage of the distance between the same samples and the class related distance. First, the class correlation distance between the training samples is calculated and stored. Second, the K nearest neighbor samples are selected, the class correlation distance of training samples and the difference ratio between the test samples and training samples are calculated respectively. Finally, according to the difference ratio, we classify the different types of samples. The experimental result shows that the algorithm combined with user rating preference can get lower MAE value, and the recommendation effect is better. With the change of K value, CCDKNN algorithm is obviously better than KNN algorithm and DWKNN algorithm, and the accuracy performance is more stable. The algorithm improves the accuracy of similarity and predictability, which has better performance than the traditional algorithm.

Robust Personalized Ranking from Implicit Feedback

2015 ◽  
Vol 30 (01) ◽  
pp. 1659001 ◽  
Author(s):  
Gai Li ◽  
Liyang Wang ◽  
Weihua Ou
Keyword(s):  
Collaborative Filtering ◽  
Nearest Neighbor ◽  
Training Data ◽  
Loss Functions ◽  
Stochastic Gradient Descent ◽  
Noise Sensitivity ◽  
Implicit Feedback ◽  
K Nearest Neighbor ◽  
Bootstrap Sampling ◽  
Personalized Ranking

In this paper, we investigate the problem of personalized ranking from implicit feedback (PRIF). It is a more common scenario (e.g. purchase history, click log and page visitation) in recommender systems. The training data are only binary in these problems, reflecting the users’ actions or inactions. One shortcoming of previous PRIF algorithms is noise sensitivity: outliers in training data might bring significant fluctuations in the training process and lead to inaccuracy of the algorithm. In this paper, we propose two robust PRIF algorithms to solve the noise sensitivity problem of existing PRIF algorithms by using the pairwise sigmoid and pairwise fidelity loss functions. These two pairwise loss functions are flexible and can easily be adopted by popular collaborative filtering models such as the matrix factorization (MF) model and the K-nearest-neighbor (KNN) model. A learning process based on stochastic gradient descent with bootstrap sampling is utilized for the optimization. Experiments are conducted on practical datasets containing noisy data points or outliers. Results demonstrate that the proposed algorithms outperform several state-of-the-art one class collaborative filtering (OCCF) algorithms on both the MF and KNN models over different evaluation metrics.

scholarly journals REKOMENDASI SISTEM PEMILIHAN MOBIL MENGGUNAKAN K-NEAREST NEIGHBOR (KNN) COLLABORATIVE FILTERING

2019 ◽  
Vol 4 (1) ◽  
pp. 26
Author(s):  
Ilham Gumantung Gusti ◽  
Muhammad Nasrun Hasibuan ◽  
Ratna Astuti Nugrahaeni
Keyword(s):  
Collaborative Filtering ◽  
Nearest Neighbor ◽  
K Nearest Neighbor

Mobil merupakan kendaraan yang sangat dibutuhkan pada masa ini. Banyak dari pengguna ketika ingin memilih mobil hanya mengetahui sebagian dari informasi mobil yang disukainya tanpa mengetahui informasi mobil lain yang sejenis. Rekomendasi sistem pemilihan mobil merupakan sistem yang dapat digunakan oleh pengguna dalam memilih mobil. Dengan diterapkannya rekomendasi sistem pemilihan mobil, pengguna akan mendapatkan informasi lebih mengenai mobil yang ingin dipilih, dan mobil lain yang mungkin mobil tersebut sama sekali belum diketahui oleh pengguna. Dalam rekomendasi sistem pemilihan mobil, penulis menerapkan metode K-Nearest Neighbor (KNN) Collaborative Filtering yang dilakukan berdasarkan jarak kedekatan Data Testing dengan Data Training. Kedekatan data (kemiripan data) tersebut digunakan untuk merekomendasikan mobil ke pengguna. Hasil yang diperoleh dalam penelitian ini adalah jika ingin mendapatkan 10 mobil terbaik maka jarak maksimal yang digunakan adalah 5%, dan akurasi terbaik didapatkan ketika K = 10 yaitu sebesar 95,15%.

scholarly journals Analisis Perbandingan Model Matrix Factorization dan K-Nearest Neighbor dalam Mesin Rekomendasi Collaborative Berbasis Prediksi Rating

2021 ◽  
Vol 5 (4) ◽  
pp. 506
Author(s):  
Janny Eka Prayogo ◽  
Aries Suharso ◽  
Adhi Rizal
Keyword(s):  
Collaborative Filtering ◽  
Matrix Factorization ◽  
Nearest Neighbor ◽  
Prediction Models ◽  
Mean Squared Error ◽  
Absolute Error ◽  
K Nearest Neighbor ◽  
Filtering Method ◽  
Factorization Model ◽  
The Matrix

Rating is a form of assessment of the likes or dislikes of a user or customer for an item. Where the higher the rating number given, the item is preferred by customers or users. In the recommendation engine, a set of ratings can be predicted and used as an object to generate a recommendation by the Collaborative Filtering method. In the Collaborative Filtering method, there is a rating prediction model, namely the Matrix Factorization and K-Nearest Neighbor models. This study analyzes the comparison of the two prediction models based on the value of Mean Absolute Error (MAE), Root Mean Squared Error (RMSE) and the prediction results generated using the movielens film rating dataset. From the analysis and testing results, it was found that MAE = 0.6371 and RMSE = 0.8305 for the Matrix Factorization model, while MAE = 0.6742 and RMSE = 0.8863 for the K-Nearest Neighbor model. The best model is Matrix Factorization because the MAE and RMSE values are lower than the K-Nearest Neighbor model and have the closest predicted rating results from the original rating value.

scholarly journals Nearby Product Recommendation System Based on Users Rating

2019 ◽  
pp. 963-968 ◽  
Author(s):  
Akanksha Jyoti ◽  
Abhijeet Roy ◽  
Suraj Singh ◽  
Nabab Shaikh ◽  
Payal Desai
Keyword(s):  
Collaborative Filtering ◽  
Euclidean Distance ◽  
Recommendation System ◽  
Nearest Neighbor ◽  
Similarity Score ◽  
The Internet ◽  
K Nearest Neighbor ◽  
Past Behavior ◽  
Rate Score ◽  
K Nearest Neighbor Algorithm

The recommendation system is very popular nowadays. Recommendation system emerged over the last decade for better findings of things over the internet. Most websites use a recommendation system for tracking and finding items by the user's behavior and preferences. Netflix, Amazon, LinkedIn, Pandora etc. platform gets 60%-70% views results from recommendation. The purpose of this paper is to introduce a recommendation system for local stores where the user gets a nearby relevant recommended item based on the rating of other local users. There are various types of recommendation systems one is User-based collaborative filtering by which the system built upon and uses user's past behavior like ratings and gives similar results made by another user. In collaborative filtering uses Euclidean distance algorithm is used to find the user's rate score to make relations with other users and Euclidean distance similarity score distinguish similarity between users. K-nearest neighbor algorithm is used to implement and find the number of users like new user where K is several similar users. Integrate with map interface to find shortest distances among stores whose product are recommended. The dataset of JSON is used to parse through the algorithm. The result shows a better approach towards the recommendation of products among local stores within a region.

scholarly journals Improving the Prediction Quality in Memory-Based Collaborative Filtering Using Categorical Features

Electronics ◽  
2021 ◽  
Vol 10 (2) ◽  
pp. 214
Author(s):  
Lei Chen ◽  
Yuyu Yuan ◽  
Jincui Yang ◽  
Ahmed Zahir
Keyword(s):  
Collaborative Filtering ◽  
Recommender Systems ◽  
Prediction Accuracy ◽  
Nearest Neighbor ◽  
Side Information ◽  
Weight Matrix ◽  
K Nearest Neighbor ◽  
Proposed Model ◽  
Real World Datasets ◽  
Better Than

Despite years of evolution of recommender systems, improving prediction accuracy remains one of the core problems among researchers and industry. It is common to use side information to bolster the accuracy of recommender systems. In this work, we focus on using item categories, specifically movie genres, to improve the prediction accuracy as well as coverage, precision, and recall. We derive the user’s taste for an item using the ratings expressed. Similarly, using the collective ratings given to an item, we identify how much each item belongs to a certain genre. These two vectors are then combined to get a user-item-weight matrix. In contrast to the similarity-based weight matrix in memory-based collaborative filtering, we use user-item-weight to make predictions. The user-item-weights can be used to explain to users why certain items have been recommended. We evaluate our proposed method using three real-world datasets. The proposed model performs significantly better than the baseline methods. In addition, we use the user-item-weight matrix to alleviate the sparsity problem associated with correlation-based similarity. In addition to that, the proposed model has a better computational complexity for making predictions than the k-nearest neighbor (kNN) method.

Sign in / Sign up

Export Citation Format

Share Document