Cross-Domain Recommender Systems

Cantador, Iván; Fernández-Tobías, Ignacio; Berkovsky, Shlomo; Cremonesi, Paolo

doi:10.1007/978-1-4899-7637-6_27

Cited by 152 publications

(80 citation statements)

References 53 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…More diverse outcomes can be obtained by gathering rapidly additional information on users' preferences. This can be achieved by actively eliciting the user to make more informative choices [31], by integrating information from other user activities [5], or by using hybrid techniques to combine recommendations obtained by different methods [24]. As an alternative to these techniques, one can leverage the additional information to construct a network and run a standard diffusion-based algorithm on it.…”

Section: Cold-start Problemsmentioning

confidence: 99%

Recommending investors for new startups by integrating network diffusion and investors’ domain preference

Zhang

Lü

2020

Information Sciences

View full text Add to dashboard Cite

Over the past decade, many startups have sprung up, which create a huge demand for financial support from venture investors. However, due to the information asymmetry between investors and companies, the financing process is usually challenging and time-consuming, especially for the startups that have not yet obtained any investment. Because of this, effective data-driven techniques to automatically match startups with potentially relevant investors would be highly desirable. Here, we analyze 34, 469 valid investment events collected from www.itjuzi.com and consider the cold-start problem of recommending investors for new startups. We address this problem by constructing different tripartite network representations of the data where nodes represent investors, companies, and companies' domains. First, we find that investors have strong domain preferences when investing, which motivates us to introduce virtual links between investors and investment domains in the tripartite network construction. Our analysis of the recommendation performance of diffusion-based algorithms applied to various network representations indicates that prospective investors for new startups are effectively revealed by integrating network diffusion processes with investors' domain preference.Our main goal is to fill this gap by designing and validating recommendation system techniques to identify suitable investors for new startups. To this end, we analyze 34,469 investment events collected from www.itjuzi.com [1]. As we focus on startups that received no previous investments in the past, widely-studied recommendation techniques based on bipartite networks [49] are not applicable here since the new startups with no previous investors turn out to be isolated nodes in the investor-company bipartite network, making them unreachable by diffusion processes. Our problem can be classified as a cold-start problem [35]: When a new actor enters the system, there is insufficient past information to provide him/her with a recommendation [26].To overcome the cold-start problem, we resort to tagging techniques [16].On the www.itjuzi.com platform, each startup is required to provide several tags that define its business scope. We found that the investors have a strong preference toward a small number of tags -in particular, in the majority of cases, an investor tends to invest in startups that feature her favorite tag. Hence, motivated by this finding and the key role played by the industry field in investment decision making [10, 28], we use tags as a key piece of information to generate recommendations.By leveraging startups' tag information, we construct three different tripartite network representations of the investment system. The most natural tripartite representation is one where investors are connected to the startups they invested in, and startups are connected to their self-reported tags -we refer to this representation as the tagcompany-investor (TCI) representation. This representation is natural because it is directly based on the collected ...

show abstract

Section: Cold-start Problemsmentioning

confidence: 99%

Recommending investors for new startups by integrating network diffusion and investors’ domain preference

Zhang

Lü

2020

Information Sciences

View full text Add to dashboard Cite

show abstract

“…This problem has been addressed from various perspectives in different research areas. It has been tackled by means of user preference aggregation and media-tion strategies for the cross-system personalization problem in user modeling [1,3,67,12], as a potential solution to mitigate the cold-start and sparsity problems in recommender systems [16,68,71,24] and as a practical application of knowledge transfer in machine learning [27,45,55].…”

Section: Cross-domain Recommender Systemsmentioning

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

Self Cite

113

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

“…Cross-domain recommender systems use multiple domains to generate recommendations, which can be categorized based on domain levels [4,5] Depending on whether overlapping occurs in the set of users or items [7], there are four situations that enable cross-domain recommendations: a) no overlap between items and users, b) user sets of different domains overlap, c) item sets overlap, and d) item and user sets overlap.…”

Section: Cross-domain Recommendationsmentioning

confidence: 99%

Improving Serendipity and Accuracy in Cross-Domain Recommender Systems

Kotkov

Wang

Veijalainen

2017

Lecture Notes in Business Information Processing

View full text Add to dashboard Cite

Abstract. Cross-domain recommender systems use information from source domains to improve recommendations in a target domain, where the term domain refers to a set of items that share attributes and/or user ratings. Most works on this topic focus on accuracy but disregard other properties of recommender systems. In this paper, we attempt to improve serendipity and accuracy in the target domain with datasets from source domains. Due to the lack of publicly available datasets, we collect datasets from two domains related to music, involving user ratings and item attributes. We then conduct experiments using collaborative filtering and content-based filtering approaches for the purpose of validation. According to our results, the source domain can improve serendipity in the target domain for both approaches. The source domain decreases accuracy for contentbased filtering and increases accuracy for collaborative filtering. The improvement of accuracy decreases with the growth of non-overlapping items in different domains.

show abstract

Cross-Domain Recommender Systems

Cited by 152 publications

References 53 publications

Recommending investors for new startups by integrating network diffusion and investors’ domain preference

Recommending investors for new startups by integrating network diffusion and investors’ domain preference

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

Improving Serendipity and Accuracy in Cross-Domain Recommender Systems

Contact Info

Product

Resources

About