Pairwise preference regression for cold-start recommendation

Park, Seung-Taek; Chu, Wei

doi:10.1145/1639714.1639720

Cited by 252 publications

(156 citation statements)

References 26 publications

(22 reference statements)

Supporting

Mentioning

155

Contrasting

Unclassified

Order By: Relevance

“…The new item problem [38,39] arises due to the fact that the new items entered in RS do not usually have initial votes, and therefore, they are not likely to be recommended. In turn, an item that is not recommended goes unnoticed by a large part of the users community, and as they are unaware of it they do not rate it; in this way, we can enter a vicious circle in which a set of items of the RS are left out of the votes/recommendations process.…”

Section: The Cold-start Issuementioning

confidence: 99%

“…When there are not enough users in particular and votes in general, it is difficult to maintain new users, which come across a RS with contents but no precise recommendations. The most common ways of tackling the problem are encouraging votes to be made via other means or not making CF-based recommendations until there are enough users and votes.The new item problem [38,39] arises due to the fact that the new items entered in RS do not usually have initial votes, and therefore, they are not likely to be recommended. In turn, an item that is not recommended goes unnoticed by a large part of the users community, and as they are unaware of it they do not rate it; in this way, we can enter a vicious circle in which a set of items of the RS are left out of the votes/recommendations process.…”

mentioning

confidence: 99%

See 1 more Smart Citation

A collaborative filtering approach to mitigate the new user cold start problem

Bobadilla

Ortega

Hernando

et al. 2012

Knowledge-Based Systems

422

202

View full text Add to dashboard Cite

In order to measure the quality of the results of a RS, there is a wide range of metrics which are used to evalúate both the prediction and recommendation quality of these systems [17,1,18,6].CF based RS estímate the valué of an item not voted by a user via the ratings made on that item by a set of similar users. The overall quality in the prediction is called accuracy [3] and the mean absolute error (MAE) is normally used to obtain it [17]. The system's ability to make estimations is called coverage and it indicates the percentage of prediction which we can make using the set of similar users selected (usually, the more similar users we select and the more votes the selected users have cast, the better the coverage we achieve). In RS, besides aiming to improve the quality measures of the predictions (accuracy and coverage), there are other issues that need be taken into account [56,41,51]: avoiding overspecialization phenomena, finding good Ítems, credibility of recommendations, precisión and recall measures, etc.The rest of the paper is divided into the following sections (with the same numbering shown here):2. State of the art, in which a review is made of the most relevant contributions that exist in the CF aspects covered in the paper: cold-start and application of neural networks to the RS. 3. General hypothesis and motivations: what we aim to contribute and the indications that lead us to believe that carrying out research into this subject will provide satisfactory results that support the hypothesis set out. 4. Design of the user cold-start similarity measure: explanation and formalization of the design of the similarity measure proposed as a linear combination of simple similarity measures, by adjusting the weights using optimization techniques based on neural networks. 5. Collaborative filtering specifications: formalization of the CF methodology which specifies the way to predict and recommend, as well as to obtain the quality valúes of the predictions and recommendations. This is the formalization that supports the design of experiments carried out in the paper. The methodology is provided which describes the use of leave-one-out cross validation applied to obtaining the MAE, coverage, precisión and recall. 6. Design of the experiments with which the quality results are obtained provided by the user cold-star similarity measure proposed and by a set of current similarity measures for which we aim to improve the results. We use the Netflix (http:// www.netflixprize.com) and Movielens (http://www.movielens.org) databases. 7. Graphical results obtained in the experiments, complemented with explanations of the behavior of each quality measure. 8. Most relevant conclusions obtained. State of the art The cold-start issueThe cold-start problem [48,1] occurs when it is not possible to make reliable recommendations due to an initial lack of ratings. We can distinguish three kinds of cold-start problems: new community, new item and new user. The last kind is the most important in RS that are already in operat...

show abstract

Section: The Cold-start Issuementioning

confidence: 99%

mentioning

confidence: 99%

A collaborative filtering approach to mitigate the new user cold start problem

Bobadilla

Ortega

Hernando

et al. 2012

Knowledge-Based Systems

422

202

View full text Add to dashboard Cite

show abstract

“…The new item problem (Park & Tuzhilin, 2008;Park & Chu, 2009) arises because the new items entered in RS do not usually have initial ratings, and therefore, they are not likely to be recommended. In turn, an item that is not recommended keeps unnoticed by a large part of the community of users, and as they are unaware of it, they do not rate it; this involves vicious circle in which a set of items of the RS are left out of the ratings/recommendations process.…”

Section: The Cold Start Problemmentioning

confidence: 99%

Incorporating group recommendations to recommender systems: Alternatives and performance

Ortega

Bobadilla

Hernando

et al. 2013

Information Processing & Management

View full text Add to dashboard Cite

Proporcionaré una formalización completa del método propuesto. Explicaré cómo obtener el conjunto de k vecinos del grupo de usuarios y mostraré cómo obtener recomendaciones usando dichos vecinos. Asimismo, incluiré un ejemplo detallando cada paso del método propuesto en un sistema de recomendación compuesto por 8 usuarios y 10 items.Las principales características del método propuesto son: (a) es más rápido (más eficiente) que las alternativas proporcionadas por otros autores, y (b) es al menos tan exacto y preciso como otras soluciones estudiadas. Para contrastar esta hipótesis realizaré varios experimentos que miden la precisión, la exactitud y el rendimiento del método. Los resultados obtenidos se compararán con los resultados de otras alternativas utilizadas en la recomendación de grupos. Los experimentos se realizarán con las bases de datos de MovieLens y Netflix. VIII IX ABSTRACTThe importance of recommender systems has grown exponentially with the advent of social networks. In this PhD thesis I will provide a wide vision about the state of the art of recommender systems. They were initially based on demographic, contentbased and collaborative filtering. Currently, these systems incorporate some social information to the recommendation process. In the future, they will use implicit, local and personal information from the Internet of Things.As we will see here, recommender systems based on collaborative filtering can be used to perform recommendations to group of users. Previous works have made this modification in different stages of the collaborative filtering algorithm: establishing the neighborhood, prediction phase and determination of recommended items. In thisPhD thesis I will provide a new method that carry out the unification process (many users to one group) in the first stage of the collaborative filtering algorithm: similarity metric computation.I will provide a full formalization of the proposed method. I will explain how to obtain the k nearest neighbors of the group of users and I will show how to get recommendations using those users. I will also include a running example of a recommender system with 8 users and 10 items detailing all the steps of the method I will present.The main highlights of the proposed method are: (a) it will be faster (more efficient) that the alternatives provided by other authors, and (b) it will be at least as precise and accurate as other studied solutions. To check this hypothesis I will conduct several experiments measuring the accuracy, the precision and the performance of my method. I will compare these results with the results generated by other methods of group recommendation. The experiments will be carried out using MovieLens and Netflix datasets. The kinds of filtering most used at the beginning of the RS (collaborative, content-based and demographic) were described in Pazzani (1999). Breese,

show abstract

“…However, regardless of the specific variant that is used, CF methods have a common limitation: the so called new user cold-start problem, which occurs when a system cannot generate personalized and relevant recommendations for a user who has just registered into the system. Although many solutions have been proposed [23,24,33,56,58,72,47,69], this problem is still challenging, and there is not a unique solution for it that can be applied to any domain or situation. Indeed, as we shall show later, different approaches better suit specific situations, e.g., when the new user has entered either zero or only a few ratings.…”

Section: Introductionmentioning

confidence: 99%

Alleviating the new user problem in collaborative filtering by exploiting personality information

Fernández-Tobías

Braunhofer

Elahi

et al. 2016

User Model User-Adap Inter

109

View full text Add to dashboard Cite

Esta es la versión de autor del artículo publicado en: This is an author produced version of a paper published in: In this paper we analyze viable solutions to the new user problem in collaborative filtering that are based on the exploitation of user personality information: (a) personality-based collaborative filtering, which directly improves the recommendation prediction model by incorporating user personality information; (b) personality-based active learning, which utilizes personality information for identifying additional useful preference data in the target recommendation domain to be elicited from the user; and (c) personality-based cross-domain recommendation, which exploits personality information to better use user preference data from auxiliary domains which can be used to compensate the lack of user preference data in the target domain. We benchmark the effectiveness of these methods on large datasets that span several domains, namely movies, music and books. Our results show that personality-aware methods achieve performance improvements that range from 6% to 94% for users completely new to the system, while increasing the novelty of the recommended items by 3% to 40% with respect to the non-personalized popularity baseline. We also discuss the limitations of our approach and the situations in which the proposed methods can be better applied, hence providing guidelines for researchers and practitioners in the field.

show abstract

Pairwise preference regression for cold-start recommendation

Cited by 252 publications

References 26 publications

A collaborative filtering approach to mitigate the new user cold start problem

A collaborative filtering approach to mitigate the new user cold start problem

Incorporating group recommendations to recommender systems: Alternatives and performance

Alleviating the new user problem in collaborative filtering by exploiting personality information

Contact Info

Product

Resources

About