A probabilistic model to resolve diversity–accuracy challenge of recommendation systems

Javari, Amin; Jalili, Mahdi

doi:10.1007/s10115-014-0779-2

Cited by 75 publications

(35 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Li et al [24] handle news variety explicitly through random walks in the user-item a nity graph created in advance. Likewise, a probabilistic model is proposed in [19] to address this aspect. In contrast, having analyzed the transition matrices extracted from the CLEF NewsREEL 2014 competition data set, Doychev et al [9] show that users tend to read news from the same category.…”

Section: News Recency Popularity and Varietymentioning

confidence: 99%

See 1 more Smart Citation

Recommending Personalized News in Short User Sessions

Epure

Kille

Ingvaldsen

et al. 2017

Proceedings of the Eleventh ACM Conference on Recommender Systems

View full text Add to dashboard Cite

News organizations employ personalized recommenders to target news articles to speci c readers and thus foster engagement. Existing approaches rely on extensive user pro les. However frequently possible, readers rarely authenticate themselves on news publishers' websites. is paper proposes an approach for such cases. It provides a basic degree of personalization while complying with the key characteristics of news recommendation including news popularity, recency, and the dynamics of reading behavior. We extend existing research on the dynamics of news reading behavior by focusing both on the progress of reading interests over time and their relations. Reading interests are considered in three levels: short-, medium-, and long-term. Combinations of these are evaluated in terms of added value to the recommendation's performance and ensured news variety. Experiments with 17-month worth of logs from a German news publisher show that most frequent relations between news reading interests are constant in time but their probabilities change. Recommendations based on combined shortterm and long-term interests result in increased accuracy while recommendations based on combined short-term and medium-term interests yield higher news variety.

show abstract

Section: News Recency Popularity and Varietymentioning

confidence: 99%

“…Traditionally, recommender solutions are of grouped into two types [1,14,19]. A content-based recommender suggests items similar to previously liked ones.…”

Section: Literature Reviewmentioning

confidence: 99%

Recommending Personalized News in Short User Sessions

Epure

Kille

Ingvaldsen

et al. 2017

Proceedings of the Eleventh ACM Conference on Recommender Systems

View full text Add to dashboard Cite

show abstract

“…In [13] introduced a probabilistic structure to resolve the diversity-accuracy dilemma in recommender systems. They proposed a hybrid model with adjustable level of diversity and precision such that one can perform this by tuning a single parameter.…”

Section: Literature Reviewmentioning

confidence: 99%

Review of Techniques for Recommender Systems

Kaur¹,

Garag²,

Kaur³

2017

IJARCSSE

View full text Add to dashboard Cite

Abstract-In electronic commerce, recommender systems are used to help customers to choose products according to their needs. These systems suggest products automatically to users by learning their requirements. Recommendations provided by these systems depends upon users purchase probability and preferences. In this paper, different techniques used for recommender systems are studied.Keywords-Recommendation Systems, E-commerce, Content Based Filtering, Collaborative Filtering, Hybrid Methods I. INTRODUCTIONRecommender systems touch our lives every day, from searching on Google to shopping on at any major online retailer. Their sophisticated algorithms attempt and often succeed at showing us the information and products we seek. The information consumers give to service providers and retailers is expanding rapidly, which makes recommendation systems both more complex, and potentially more powerful. Online behaviour including customer metadata, transaction histories and communications allows companies to understand shoppers better, sense their similarities, and address needs they may not even know they have. Simultaneously, we are better able to analyse information about the items sought, including images, sounds in the case of music, and descriptions, which allows companies to refine the ways they cluster products, as well as whom they target as potential buyers. Recommendation systems are at the center of retail both online and off, showing ads on web sites as well as through dynamic displays in brick and mortar stores. By using facial recognition and in-store video cameras, retailers are able to group their customers instantly by gender, age and other demographics in order to show them immediately the goods they are most likely to buy. Given the explosion of consumer goods as well as the rapid increase in the number of vendors on the Internet, the main problem facing customers is how to find the object that they seek, when it is buried beneath a mountain of irrelevant information. A recommender system that can instantly personalize ads in order to solve the customer's search problem is doing both the customer, and the vendor, a favour. New goods are constantly introduced, and new fads sweep the nation, altering shoppers' behaviour. Only a recommender system that is constantly able to learn new patterns can serve consumer needs.

show abstract

“…Up to now, several model-based methods have been proposed in the literature. These methods include probabilistic models [19,21,22], clustering models [8,13,23,24], dimensionality reduction techniques [25,26], pattern mining approach [14], latent semantic models [20], and Markov decision process-based CF systems [22,27]. A review of research indicates that memory-based techniques are widely used as compared to model-based techniques [10,[28][29][30].…”

Section: Introductionmentioning

confidence: 99%