An efficient hybrid similarity measure based on user interests for recommender systems

Hawashin, Bilal; Lafi, Mohammad; Kanan, Tarek; Mansour, Ayman

doi:10.1111/exsy.12471

Cited by 31 publications

(15 citation statements)

References 40 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Early approaches in recommender systems were based on the popular content‐based filtering (CBF) algorithms. These algorithms model user profiles by associating their preferences with the item content (Degemmis, Lops, & Semeraro, 2007; Eirinaki, Vazirgiannis, & Varlamis, 2003; Hawashin, Lafi, Kanan, & Mansour, 2019; Jannach et al, 2010; Magnini & Strapparava, 2001; Martins, Belém, Almeida, & Gonçalves, 2016; Renckes, Polat, & Oysal, 2012). The user preferences can be of different forms, for example, ratings or interactions, and can be elicited explicitly (Billsus & Pazzani, 1999), or implicitly (Kelly & Teevan, 2003).…”

Section: Related Workmentioning

confidence: 99%

Addressing the New Item problem in video recommender systems by incorporation of visual features with restricted Boltzmann machines

Hazrati

Elahi

2020

Expert Systems

View full text Add to dashboard Cite

Over the past years, the research of video recommender systems (RSs) has been mainly focussed on the development of novel algorithms. Although beneficial, still any algorithm may fail to recommend video items that the system has no form of data associated to them (New Item Cold Start). This problem occurs when a new item is added to the catalogue of the system and no data are available for that item. In content‐based RSs, the video items are typically represented by semantic attributes, when generating recommendations. These attributes require a group of experts or users for annotation, and still, the generated recommendations might not capture a complete picture of the users' preferences, for example, the visual tastes of users on video style. This article addresses this problem by proposing recommendation based on novel visual features that do not require human annotation and can represent visual aspects of video items. We have designed a novel evaluation methodology considering three realistic scenarios, that is, (a) extreme cold start, (b) moderate cold start and (c) warm‐start scenario. We have conducted a set of comprehensive experiments, and our results have shown the superior performance of recommendations based on visual features, in all of the evaluation scenarios.

show abstract

Section: Related Workmentioning

confidence: 99%

Addressing the New Item problem in video recommender systems by incorporation of visual features with restricted Boltzmann machines

Hazrati

Elahi

2020

Expert Systems

View full text Add to dashboard Cite

show abstract

“…TF-IDF also used for several recommender system [10,11,12,13,14]. One research also propose a recommendation system for academic collaboration in undergraduate research [15] based on the undergraduates' profiles and incorporates rule-based filtering techniques. A decision support system by applying the simple additive weighting method proposed by Pristiwanto [16] for the determination of the final assignment supervisor also provides similar aim.…”

Section: Literature Reviewmentioning

confidence: 99%

Research Supervisor Recommendation System Based on Topic Conformity

Rismanto¹,

Syulistyo²,

Agusta³

2020

IJMECS

View full text Add to dashboard Cite

In a higher education such as universities, final project are under supervision of one or more supervisors with a similar research interest or topic. The determination of the final project supervisor is an important factor in the work of the student's final project. However, the lack of information about the supervisor can hamper students in making the determination of the supervisor. Thus, a system is needed that can facilitate students in determining the final project or thesis advisors in accordance with the research topic. This problem is the basis of this research. The study is conducted by developing a web-based system and applying the TF-IDF word weighting and cosine similarity method. TF-IDF method is a way to give the weight of the relationship of a word to the document. The cosine similarity is a method for calculating the similarity between two objects expressed in two vectors by using keywords from a document as a measure. The results of the advisor recommendation system can provide recommendations to students regarding the final assignment advisor who has conducted research in accordance with the topic of the student's final assignment written in Indonesian. In 20 testing, the accuracy of the comparison of the results of the system recommendations with the actual data obtained an average of 75% by comparing system recommendation with actual assigned supervisor.

show abstract

“…The results reveal that although this method did not increase the global coverage, it improved the predictions of already covered items, alleviating some of the drawbacks of the sparsity problem. Hawashin et al [38] proposed a hybrid similarity measure based on explicit user interests; the proposed method achieved good performance even when no co-rated items existed between two users. However, it did not consider the semantic meanings of the ratings, and it depended on the existence and quality of explicit user interests.…”

Section: Sparsity In Neighborhood-based Modelsmentioning

confidence: 99%

Switching Hybrid Method Based on User Similarity and Global Statistics for Collaborative Filtering

Ortal

Edahiro

2020

IEEE Access

View full text Add to dashboard Cite

Collaborative filtering (CF) is a technique used in recommender systems to provide meaningful suggestions based on known feedback obtained from like-minded users. The measure of similarity plays a critical role in the performance of neighborhood-based CF methods. However, conventional similarity measures suffer from limitations because they only consider the direction of the rating vectors. We propose a novel similarity measure that considers the semantic nuances of the ratings; in particular, it weights the contributions of ratings in proportion to the users' degree of indifference towards the items. Additionally, to address the sparsity problem that affects the performance of CF techniques, we propose a switching hybrid method that predicts user ratings based on either our custom similarity measure or through user and item biases. We evaluated the proposed method on six different datasets and compared it with other CF methods. The results show that the proposed recommender consistently outperforms those using conventional similarity measures when the sparsity of the dataset is high.

show abstract

An efficient hybrid similarity measure based on user interests for recommender systems

Cited by 31 publications

References 40 publications

Addressing the New Item problem in video recommender systems by incorporation of visual features with restricted Boltzmann machines

Addressing the New Item problem in video recommender systems by incorporation of visual features with restricted Boltzmann machines

Research Supervisor Recommendation System Based on Topic Conformity

Switching Hybrid Method Based on User Similarity and Global Statistics for Collaborative Filtering

Contact Info

Product

Resources

About