Performance analysis of recommendation system based on collaborative filtering and demographics

Kapetanakis

et al. 2019

Vietnam J. Comp. Sci.

Recommender systems help users find relevant items efficiently based on their interests and historical interactions with other users. They are beneficial to businesses by promoting the sale of products and to user by reducing the search burden. Recommender systems can be developed by employing different approaches, including collaborative filtering (CF), demographic filtering (DF), content-based filtering (CBF) and knowledge-based filtering (KBF). However, large amounts of data can produce recommendations that are limited in accuracy because of diversity and sparsity issues. In this paper, we propose a novel hybrid method that combines user–user CF with the attributes of DF to indicate the nearest users, and compare four classifiers against each other. This method has been developed through an investigation of ways to reduce the errors in rating predictions based on users’ past interactions, which leads to improved prediction accuracy in all four classification algorithms. We applied a feature combination method that improves the prediction accuracy and to test our approach, we ran an offline evaluation using the 1M MovieLens dataset, well-known evaluation metrics and comparisons between methods with the results validating our proposed method.

Section: Hybrid Recommendation Approachesmentioning

confidence: 99%

Improved Movie Recommendations Based on a Hybrid Feature Combination Method

Kapetanakis

et al. 2019

Vietnam J. Comp. Sci.

“…Demographic Recommender system [34] considers demographic data in collaborative filtering when providing recommendations. In [35], recommendations are generated based on the product demographic data learned from online www.ijacsa.thesai.org reviews and blog entries.…”

Section: E Demographic Based Recommender Systemmentioning

confidence: 99%

Personalized Recommender by Exploiting Domain based Expert for Enhancing Collaborative Filtering Algorithm :PReC

Sridevi¹,

Rao²

2019

ijacsa

The large amount of information available on the internet initiated various Recommender algorithms to act as an intermediate between number of choices and internet users. Collaborative filtering is one of the most traditional and intensively used recommendation approaches for many commercial services. Despite providing satisfying outcomes, it does have some issues that include source diversity, reliability, sparsity of data, scalability and cold start. Thus, there is a need for further improvement in the current generation of recommender system to achieve a more effective human decision support in a wide variety of applications and scenarios. Personalized Expert based collaborative filtering (PReC) approach is proposed to identify domain specific experts and the use of experts preference enhanced the performance of collaborative filtering recommender systems. A unified framework is proposed that integrates similar users rating data, experts rating and demographic data to reduce the number of pairwise computations from the search space to ensure scalability and enabled fine grained recommendations. The proposed method is evaluated using accuracy metrics MAE, RMSE on the data set collected from MovieLens datasets.

“…Adjusted cosine similarity is also a similarity measure which is used in collaborative filtering based recommender system. It is used in the case in which difference in every user's use of rating scale is considered [8]…”

Section: Adjusted Cosine Similaritymentioning

confidence: 99%

“…In [3] K-Nearest-Neighbor (KNN) classification is employed to be used on-line to spot clients/visitors click stream knowledge, matching it to a specific user group and advocate a tailored browsing choice that meet the necessity of the precise user at a selected time. They stated that the K-NN classifier is clear, consistent, simple, easy to know, high affinity to have desirable qualities and straightforward to implement than most alternative machine learning algorithms specifically once there is very little or no previous information regarding data distribution.In [8], item ratings from item based collaborative filtering recommender techniques are associated with ratings computed from user clusters based on demographics in a weighted manner. The stated solution is scalable and successfully overcome user based cold start.…”

Section: Adjusted Cosine Similaritymentioning

confidence: 99%

An Approach for Recommender System by Combining Collaborative Filtering with User Demographics and Items Genres

Tiwari¹,

Shrivastava²

2015

IJCA

With the explosion of service based web application like online news, shopping, bidding, libraries great amount of information is available. Due to this information overload problem, to find right thing is a tedious task for the user. A recommender system can be used to suggest customized information according to user preferences Collaborative filtering techniques play a vital role in designing the recommendation systems. The collaborative filtering technique based recommender system may suffer with cold start problem i.e. new user problem and new item problem and scalability issues. Traditional K-Nearest Neighbor Technique also suffers with user and item cold start problem.In this paper recommender system generates suggestions for user by combining collaborating filtering on transaction data with rating predicted with user demographics and item similarity. The final rating is weighted sum of ratings computed from transaction data, user data and item data. The advantage of proposed system that recommender system can deal with cold start in case of "new user" or "new item" .and Also system has low MAE and RMSE in comparison of traditional collaborative filtering based on K-Nearest Neighbor approach.