A Literature Review of Recommender Systems for the Cultural

Sector

Nguyen Anh Khoa Dam

and Thang Le Dinh

Department of Marketing and Information Systems, Université du Québec à Trois-Rivières, Trois-Rivières, Canada

Keywords: Recommender System, Cultural Sector, Business Analytics, Artificial Intelligence, SMEs/SMOs.

Abstract: Nowadays, organizations in the cultural sector have faced the problem of improving the discoverability of

their products to meet the target objective regardless of the tremendous amount of information. In this respect,

recommender systems have been proven to be the solution for enterprises, especially for cultural small and

medium-sized organizations and enterprises (SMOs/SMEs), to enhance the discoverability of their products.

This study aims at presenting a concept-centric literature review of recommender systems for cultural

SMEs/SMOs to identify the current status-quo of the application in six cultural domains, including heritage

and libraries, live performance, visual and applied arts, written and published works, audio-visual and

interactive media, and sound recording. The finding of this paper reveals the adoption of recommender

systems of cultural SMOs/SMEs is still in the early stage of maturity. The specific status-quo of recommender

system adoption in each cultural domain is uncovered through the literature review. Other relevant aspects,

which relate to data sources, data mining models, and algorithms, are also discussed in detail. Finally, the

paper proposes future research directions to promote the application of artificial intelligence in general, and

recommender systems, in particular, in the cultural sector.

1 INTRODUCTION

In the age of big data, the challenge of customers has

changed from information shortage into information

overload (Sassi, Mellouli, & Yahia, 2017). This

stimulates the need of filtering out misleading

information and efficiently searching for the right

information (Ekstrand, Riedl, & Konstan, 2011;

Schafer, Frankowski, Herlocker, & Sen, 2007; Wedel

& Kannan, 2016). Accordingly, organizations and

enterprises in the cultural sector face the problem of

improving the discoverability of their products and

services; hereafter called cultural products, to meet

target profits regardless of the tremendous amount of

information. The discoverability (Dasgupta et al.,

2007) is defined as the ability of cultural products to

be found by customers who search for them and to be

recommended to those who are unaware of them.

In this respect, recommender systems (RSs) have

been proven to be the solution for the cultural sector,

especially for small and medium-sized organizations

(SMOs) and enterprises (SMEs), to enhance the

discoverability of their products (Bartolini et al.,

https://orcid.org/0000-0003-0928-8402

https://orcid.org/0000-0002-5324-2746

2016; Kabassi, 2013; A. Moreno, Valls, Isern, Marin,

& Borràs, 2013). However, choosing and developing

the right approach for RSs seem to be a challenge for

most enterprises, especially in the case of SMOs and

SMEs with limited resources (Ekstrand et al., 2011;

Wedel & Kannan, 2016).

For this reason, this study aims at presenting a

literature review on recommender systems in the

cultural sector for SMEs/SMOs. Therefore, the

primary objective of the study is to identify the

current status-quo of the RS adoption in different

domains of the cultural sector. Based on the primary

objective, the second objective is to identify the

relevant aspects, which need to be taken into account,

relates to data sources, data mining models, and

algorithms for recommender systems.

This paper is structured as follows. At first, the

paper continues with the theoretical background. The

research design, including the research process and

classification framework, is also presented as the

foundation to categorize articles in this literature

review. Then the paper analyzes the reviewed articles

with an in-depth discussion. Future research

Dam, N. and Dinh, T.

A Literature Review of Recommender Systems for the Cultural Sector.

DOI: 10.5220/0009337807150726

In Proceedings of the 22nd International Conference on Enterprise Information Systems (ICEIS 2020) - Volume 1, pages 715-726

ISBN: 978-989-758-423-7

715

directions are inferred from the discussion. Finally,

the paper concludes with theoretical and practical

contributions concerning this research stream.

2 BACKGROUND

The theoretical background begins with the different

types of RSs, continues with the relations between

RSs and customer experience and satisfaction, and

ends with the adoption of RSs in the cultural sector.

Types of RSs. Recommender systems can be

classified into three types: Collaborative filtering

(CF), Content-based filtering (CB), and Hybrid

system (Felício et al., 2016; Garcia, Sebastia, &

Onaindia, 2011). A CF recommender system requires

data on users’ evaluation of purchase history to make

suggestions (Lin, 2014). In contrast, CB

recommender systems connect profiles of users'

preferences with descriptions of relevant items

(Mathew, Kuriakose, & Hegde, 2016). The hybrid

systems integrate CF techniques with CB techniques

or even with other techniques to optimize

recommendations (Li, Xu, Wan, & Sun, 2018).

RSs and Customer Experience. Recommender

systems play an important role in optimizing

customer experience from finding to engaging with

the products (Albadvi & Shahbazi, 2009; Barragáns-

Martínez et al., 2010). The most significant function

of RSs is to predict pertinent services or products

based on users’ preferences (Lu, Wu, Mao, Wang, &

Zhang, 2015; H.-R. Zhang & Min, 2016). To put it

concisely, RSs have two key functions: to predict and

to recommend the most relevant service or products

(Ekstrand et al., 2011; Schafer et al., 2007).

RSs and Customer Satisfaction. Recommender

systems have been proven as a means to increase

customer satisfaction, maintain a long-term

relationship, and improve financial performance

(Briguez et al., 2014; Siu, Zhang, Dong, & Kwan,

2013). Particularly, the applications of RSs produce

prosperous results in the field of cultural heritage,

tourism, and leisure activities (Bartolini et al., 2016;

Kabassi, 2013; A. Moreno et al., 2013). Movie and

music RSs built by Netflix and Yahoo have

noticeably improved the financial performance of

these enterprises (Christensen & Schiaffino, 2011;

Jannach, Resnick, Tuzhilin, & Zanker, 2016).

RS Adoption in the Cultural Sector. Regardless of

a wide range of adoption of RSs in the cultural sector,

studies on such research stream scatter all over the

literature. As a consequence, it is challenging to take

an overview of the big picture of the RS adoption

among different cultural domains. Even though CF

systems are proven to be effective for the

entertainment and cultural domain, they raise issues

related to user engagement and accuracy as

recommendations inferred from a community may

not be precise for an individual member (Kabassi,

2013; Villegas, Sánchez, Díaz-Cely, & Tamura,

2018). Another drawback of the CF approach is the

insufficient amount of data input, which is defined as

the cold start, impedes many SMEs/SMOs at the

early stage of adoption (Kabassi, 2013; Park, Kim,

Choi, & Kim, 2012). On the other hand, CB systems

face the problem of identifying algorithms for

effectively matching common attributes between

users and items (F. Deng, Ren, Qin, Huang, & Qin,

2018; Yao, Sheng, Segev, & Yu, 2013). The hybrid

approach combing CF and CB has shown to

outperform (van Capelleveen, Amrit, Yazan, & Zijm,

2019); yet, it requires a certain level of investment for

information technology (IT) infrastructure which can

be an obstacle for most SMEs/SMOs with financial

constraints (Barragáns-Martínez et al., 2010; Park et

al., 2012; Yao et al., 2013).

3 RESEARCH DESIGN

3.1 Research Process

As the nature of research on recommender systems

and the cultural sector spreads over various databases,

this paper builds its literature review from a wide

range of academic sources such as Science Direct,

Emerald, Business Source Premier, EBSCOhost,

ProQuest, Google Scholars, and IEEE/IEE Library

(Park et al., 2012). The search process is conducted

through different descriptors: “recommender

systems”, “recommendation systems”, “cultural

domain”, “cultural sector”, “cultural heritage”,

“movie recommender”, “music recommend*”, “book

recommend*”, “image recommend*”. Then more

than 500 articles were screened based on abstracts,

structure, and content. As the paper focuses on

SMEs/SMOs, articles with relevant content are

primarily chosen. Considering the limited number of

articles for cultural SMEs/SMOs, the search was

expanded to cultural organizations in general to

ensure the broad literature review. However, the

content of 69 selected articles is screened with the

attention on the data mining models and IT

infrastructure, which are applicable for SMEs/SMOs.

These selected articles are from top Management and

Information Systems journals such as Expert Systems

ICEIS 2020 - 22nd International Conference on Enterprise Information Systems

716

with Applications, Decision Support Systems,

Knowledge-Based System, Decision Support

Systems, IEEE, Information Sciences. These journals

are also listed in the Scimago Journal & Country

Rank in 2018.

Table 1: Distribution of articles by journal titles.

Journal Title No %

Expert Systems with Applications

IEEE

Information Science

Knowledge-Based System

Decision Support Systems

Neuro Computing

Others

24.6 %

15.9%

4.3%

8.7%

5.8%

2.9%

37.7%

Total 69 100%

According to Table 1, the number of relevant

articles comes from Expert Systems with

Applications (24.6%), followed by IEEE (15.9%) and

Knowledge-Based System (8.7%). During the

searching process, the forward and backward search

techniques are implemented to ensure that the review

of 69 chosen articles can represent the literature of

this domain. Therefore, the paper can demonstrate its

validity and reliability as a literature review.

The 69 reviewed articles are also categorized by

years of publication (Figure 1). The period of a

decade is chosen to examine the status quo and trends

of research in this field. It is noticeable that the

number of articles considerably increases compared

to the past 10 years. There is a substantial increase in

2016 and 2018. This indicates the interests of scholars

in the application of RSs in the cultural sector.

Figure 1: Distribution of articles by years of publication.

3.2 Classification Framework

The classification framework for recommender

systems in the cultural sector consists of different

domains and types of recommender systems. This

study is based on the previous study of Park et al.

(2012) whose objective is also to find out the trend of

RS adoption in various application fields.

Accordingly, the paper classifies reviewed articles

into six cultural domains and three types of

recommender systems. The classification of the

literature review is introduced in Figure 2. A detailed

description of the framework is presented in the next

part of this paper.

Figure 2: A classification framework.

Classification of Cultural Domains. Regarding the

classification of the cultural domains, the paper

classifies research articles based on the six domains

of the cultural sector according to the Canadian

Framework for Culture Statistics

(Canada, 2011).

These domains are Heritage and libraries, Live

performance, Visual and applied arts, Written and

published works, Audio-visual and interactive media,

and Sound recording. The Heritage and libraries

domain consists of definable related industries,

products, and occupations associated with archives,

libraries, cultural heritage, and natural heritage. The

Live performance domain covers activities

concerning the performing arts of theatre, artists, and

multidisciplinary events (e.g.: celebrations and

festivals). The visual and applied arts domain

contains the following sub-domains: original visual

art (e.g.: paintings, sculptures), art reproductions

(copies of original visual arts), photography, and

crafts. The written and published works domain

symbolizes different kinds of publications, including

books, periodicals, newspapers, and other published

works. The audio-visual and interactive media

domain is divided into three categories: film and

video, broadcasting (e.g.: radio, television, the

Internet) and interactive media (e.g.: online games).

The Sound recording domain represents sound

recording and music publishing.

Various RSs have been developed in different

domains of the cultural sector. The cultural heritage

domain has witnessed the evolution of various

versions of RSs in enhancing access to museums,

galleries, and other historic places. In the domains of

movies, music, and books, several RSs were built to

help users find relevant items. However, it is

A Literature Review of Recommender Systems for the Cultural Sector

717

challenging to find studies, which discuss RSs for

several different domains of the cultural sector.

Classification of RSs. As mentioned in previous

sections, recommender systems can be classified

generally into three specific types: collaborative

filtering, content-based filtering, and hybrid system.

Classifying the selected articles with types of

recommender systems indicates the current status and

trends of applications of such systems in different

domains of the cultural sector. The identification of

RS types also gives an idea of algorithms as well as

required IT architecture for developing and exploiting

the system. Therefore, it is significant to classify the

literature review based on the three types of

recommender systems. A more detailed description of

the three RSs is also presented as follows:

(1) Collaborative filtering (CF) recommender

systems are based on users’ evaluation of a set of

items such as songs, movies, images (Barragáns-

Martínez et al., 2010; Schafer et al., 2007). Under this

approach, users who share similar opinions on

specific items tend to have the same idea on other

issues (Barragáns-Martínez et al., 2010; Bobadilla,

Ortega, Hernando, & Gutiérrez, 2013). Collaborative

filtering recommender systems make predictions to a

specific user regarding the fact that these systems use

information extracted from many users (Schafer et al.,

2007; Wedel & Kannan, 2016).

(2) Content-based filtering (CB) recommender

systems match profiles of users’ preferences with

descriptions of relevant items (Barragáns-Martínez et

al., 2010; Yao et al., 2013). In these systems,

algorithms recommend similar items that a user used

to like or interact with. Three components of content-

based filtering recommender systems are profiles of

users' preferences, item descriptions, and item

comparison to user profiles (Barragáns-Martínez et

al., 2010). Content-based filtering recommender

systems seem to prevail today due to the vast content

on social media (Barragáns-Martínez et al., 2010;

Bobadilla et al., 2013).

(3) Hybrid recommender systems combine

content-based filtering with collaborative filtering or

other approaches to take advantage of each technique

(Barragáns-Martínez et al., 2010; Bobadilla et al.,

2013). The limitation of CF recommender systems

related to the lack of information about user’s

behaviors can be solved by the functionality of the CB

approach (Albanese, d'Acierno, Moscato, Persia, &

Picariello, 2011). On the contrary, the issue faced CB

recommender systems are over-specialization, which

only focuses on similar items to user preferences and

ignores "different" items although a user might want

to try something new (Borràs, Moreno, & Valls,

2014). As a matter of fact, the functionality of the CF

approach can overcome this limitation (Kazienko &

Kolodziejski, 2006). The hybrid approach is proved

to outperform the CF and CB ones (Yao et al., 2013).

4 LITERATURE REVIEW

Articles are categorized based on the classification

framework for recommender systems in the cultural

sector (Table 2). The degree of coverage of all the

articles related to each sub-category is noted with

three levels: highly covered (***), moderately

covered (**), and slightly covered (*) (Rickenberg,

Neumann, Hohler, & Breitner, 2012).

The reviewed articles are ordered in a concept-

centric way instead of being arranged based on author

names or year of publication (Webster & Watson,

2002). Accordingly, there are two key concepts:

cultural domains and RS types. On the horizontal

dimension of Table 2, the concepts are broken down

into units of analysis (Webster & Watson, 2002). The

“cultural domains” concept consists of six units of

analysis whereas the “RS types” concept covers three

units. On the vertical dimension of Table 2, the

cluster analysis approach (Rickenberg et al., 2012) is

applied to group similar articles belonging to the

same cultural domain. Reviewing the paper through

vertical clusters with horizontal units of analysis

reveals the relationships and coherency among

articles.

Cultural Domains. The number of reviewed articles

scatters over the six cultural domains as follows:

heritage and libraries (23), live performance (3),

visual and applied arts (11), written and published

works (18), audio-visual and interactive media (30),

and sound recording (28). Each cultural domain is

considered as a unit of analysis in which the degree

of intensity is calculated as the weighted total based

on the extent of coverage such as highly covered

(***x3), moderately covered (**x2), and slightly

covered (*x1) (Rickenberg et al., 2012). Compared to

the total score – the total number of articles discussing

the topic, the weighted total will better inform the

significance and trends of each unit of analysis.

Based on Table 2, recommender systems catch

the most attention in the domain of audio-visual and

interactive media with the highest intensity (70). The

application of recommender systems in the domain of

heritage and libraries keeps the second place with the

level of high intensity (59). However, the number of

articles discussing this domain (23) is less than those

in the domain of music (28). Only a small portion of

ICEIS 2020 - 22nd International Conference on Enterprise Information Systems

718

the reviewed articles deals with recommender

systems in the domains of visual and applied arts (11)

and written and published works (18). The literature

scarcely discusses the application of RSs for live

performance, including performing art, festivals, and

celebrations.

Table 2: Classification of the most relevant articles.

Cultural Domains RS Types

Heritage & libraries

Live performance

Visual &

applied arts

Works

Media

Sound recordings

Collaborative

filtering

Content-based

Hybrid

Albanese et al. (2011a) *** * * ***

Chianese et al. (2013) *** ***

Albanese et al. (2011b) *** ** ** ** ***

Borràs et al. (2014) *** *** *** ***

Bartolini et al. (2013) *** ***

Bartolini et al. (2016) *** * * ***

Kabassi (2013) *** ***

Moreno et al. (2013) *** ** ** ***

Cuomo et al. (2015) *** **

Chianese and Piccialli (2016) *** ***

Garcia et al. (2011) *** * ***

Siu et al. (2013) *** * * * * *

Portugal et al. (2018) * * * * * * ** ** **

Villegas et al. (2018) * * ** ** ** ** **

Champiri et al. (2015) *** * ** *** *** ***

Noguera et al. (2012) *** ** * ***

Park et al. (2012) * * * * * * ** ** *

Umanets et al. (2014) *** * * ***

Yang and Hwang (2013) *** * ***

Ardissono et al. (2012) *** ** ** **

Sassi et al. (2017) *** * ** ** ** ** **

Deng et al. (2018) *** ** ** ***

Sanchez et al. (2012) *** *** *** ***

Zhang et al. (2017) *** ***

Sejal et al. (2016) *** ***

Felício et al. (2016) *** ***

Tkalcic et al. (2012) *** ***

Lin (2014) ** * * ***

Albadvi and Shahbazi (2009) ** ** * * ***

Bedi and Agarwal (2011) *** ***

Bedi and Vashisth (2014) *** ***

Zhou (2010) *** ***

Hariadi and Nurjanah (2017) *** ***

Alharthi et al. (2018) *** * *** ***

Núñez-Valdéz et al. (2012) *** * * *

Nirwan et al. (2016) *** ***

Mathew et al. (2016) *** *** *** ***

A Literature Review of Recommender Systems for the Cultural Sector

719

Table 2: Classification of the most relevant articles (cont.).

Cultural Domains RS Types

Heritage & libraries

Live performance

Visual &

applied arts

Works

Media

Sound recordings

Collaborative filtering

Content-based

Hybrid

Colombo-Mendoza et al. (2015) *** ***

Katarya and Verma (2017) *** *** ***

Li et al. (2018) *** ** ***

Koren et al. (2009) *** * *** *

Ekstrand et al. (2011) *** * *** * *

Kim et al. (2011a) *** *** * *

Carrer-Neto et al. (2012) *** * * ***

Eirinaki et al. (2018) * * ** * *** * *

Beel et al. (2013) ** *

Kim et al. (2011b) * * *** *

van Capelleveen et al. (2019) * ***

Barragáns-Martínez et al. (2010) *** * ** ***

Bobadilla et al. (2013) *** ** ** ** ***

Jannach et al. (2016) *** * ** *

Hu et al. (2019) *** ***

Moreno et al. (2016) *** *** ***

Briguez et al. (2014) *** ***

Zhang and Min (2016) *** * ***

Pereira and Hruschka (2015) *** ** ***

Viktoratos et al. (2018) ** * * ***

Hyung et al. (2014) *** ***

Lee et al. (2017) *** ***

Lu et al. (2015) * * * *** *** ** ** ***

Christensen and Schiaffino (2011) *** *** * ***

Bauer and Nanopoulos (2014) * ** *** * *

Kaminskas and Ricci (2012) *** ** ** ***

Andjelkovic et al. (2019) *** ***

Zheng et al. (2018) *** *** ***

Sánchez-Moreno et al. (2016) *** ***

Deng et al. (2015) *** *** ***

Horsburgh et al. (2015) *** ***

Liu and Chen (2018) *** ***

Total

23 3 11 18 30 28 46 35 55

Weighted total:

59 3 24 36 70 55 102 62 148

Notation:

Highly covered: ***x3; Moderately covered: **x2; Slightly covered: *x1

Types of RSs. The number of articles covering each

unit of analysis, including collaborative filtering (46),

content-based filtering (35), and hybrid (55) points

out interesting facts. Accordingly, the hybrid

approach receives the most attention from scholars

with the highest intensity (148). This finding is also

supported by previous studies (Umanets et al., 2014;

van Capelleveen et al., 2019). Followed by the hybrid

ICEIS 2020 - 22nd International Conference on Enterprise Information Systems

720

approach is the collaborative filtering RS with a high

intensity of 102. The content-based RSs catches the

least attention of scholars with the low intensity (62).

5 DISCUSSION

This paper reviewed 69 academic articles to gain a

comprehensive and in-depth insight into the current

status of the applications of recommender systems in

the cultural sector. The number of reviewed articles is

selected from a broad range of scientific journals to

demonstrate a multifaceted spectrum of scholars and

opinions in this research field. Based on the research

objectives, the discussion focuses on the application

of RSs in the cultural domains and relevant aspects

related to data sources, models and algorithms.

5.1 Applications of RSs

Even though the topic of RSs is traditional in the

information system research; however, the

application of RSs is an up-to-date research stream -

especially in the cultural sector.

Audio-visual and Interactive Media, and Sound

Recording Domains. The finding of this paper

indicates that the applications of RSs dominate in the

domain of audio-visual and interactive media,

followed by the sound recording domain. This finding

reaches an agreement with the previous studies

regarding the blossom of many movies and music

RSs (Andjelkovic et al., 2019; Katarya & Verma,

2017). Netflix’s recommender system would be a

salient example as it contributes more than 75% of the

total movie downloads and rentals (Jannach et al.,

2016). The most popular music recommender system

with a significant impact on the music industry is

Last.fm (Horsburgh et al., 2015; Zheng et al., 2018).

The common goal of RSs in the movie and music RSs

is to navigate and discover songs or movies and then

share them with a specific user community (Eirinaki

et al., 2018; Kaminskas & Ricci, 2012). Therefore,

RSs in these two domains emphasize the significance

of social networks (Carrer-Neto et al., 2012).

Accordingly, hybrid RSs integrated with user-

generated tags for social RSs seem to dominate these

domains as they outperform CF and CB systems

(Kim, Alkhaldi, et al., 2011; Pereira & Hruschka,

2015). Merging various techniques, hybrid RSs can

improve user satisfaction and outperform other

systems (Andjelkovic et al., 2019; Christensen &

Schiaffino, 2011).

Heritage and Libraries Domain. In light of the

heritage and libraries domain, the application of RSs

serves two purposes: to recommend a destination for

visits and to suggest a list of activities for visitors to

enjoy such a destination (Garcia et al., 2011). To

facilitate these purposes, RSs in this cultural domain

often apply information technologies - for example, a

mobile platform - to support on-site

recommendations while users at those cultural sites

can know what to do or where to go (Ardissono et al.,

2012; Chianese et al., 2013). Location and activities-

based RSs are mostly developed upon the content-

based approach (Villegas et al., 2018). However, a

user may not satisfy with all recommendations in

which he or she shows an interest (Yang & Hwang,

2013). Therefore, collaborative filtering techniques

are integrated to form a hybrid system. Hybrid RSs

which combine multiple filtering techniques are the

preferred choice in this domain (Umanets et al.,

2014; van Capelleveen et al., 2019).

Live Performance Domain. Although the

application of RSs in the live performance domain

serves the same purposes as the heritage and libraries

domain, very few RSs are developed and

implemented. This can be explained by the occasional

and short-term characteristics of performance,

festivals, and celebrations. In this domain, investment

in RSs may not yield profitable outcomes.

Written and Published Works Domain. RSs in this

domain give prominence to the importance of trust for

reliable recommendations (Bedi & Agarwal, 2011).

Especially through the platform of Web and social

networks, recommendations from users' opinions are

rated by other users to ensure trustworthiness and

avoid scams (Bedi & Vashisth, 2014; Zhou, 2010).

The hybrid approach still dominates in developing

RSs for books, periodicals, and newspapers (Hariadi

& Nurjanah, 2017). However, it is noticeable that a

significant number of book recommendations are

based on the CB approach (Zhou, 2010). This is

rationalized by the fact that the CB can avoid the cold

start’s problem of lacking data on new items from the

CF approach (Alharthi et al., 2018). CB systems do

not rely on user ratings as they make

recommendations based on metadata of items.

Metadata reveals attributes of items related to unique

identifiers, years of publication, description, legal

aspects, enrichment, technicality, contextual or

demographic information (Alharthi et al., 2018).

Structured and standardized metadata will facilitate

the discoverability of cultural contents via RSs

(Andjelkovic et al., 2019; Eirinaki et al., 2018).

A Literature Review of Recommender Systems for the Cultural Sector

721

Visual and Applied Arts Domain. Only a small

proportion of the literature view discusses the

applications of RSs in the domain of visual and

applied arts. The hybrid approach, which is

considered the most optimal can improve up to 90%

of recommendation quality (F. Deng et al., 2018;

Tkalcic et al., 2012). Hybrid RSs take advantage of

data on users' visual perception from the content-

based approach to overcome the problem of

recommendation reliability of collaborative filtering

techniques (Felício et al., 2016). Metadata on users’

visual perception will function as a source of

contextual information (Tkalcic et al., 2012);

therefore, the problem of different degrees of affinity

among users with similar attributes can be handled

(Felício et al., 2016; Sanchez et al., 2012). The

challenges of aesthetics taste dispersion and

undefined visual tastes of content-based techniques

can also be overcome through recommendation

techniques inferred from a user community of the CF

approach (Sanchez et al., 2012).

5.2 Data Sources

Recommender systems are considered as the

significant application of knowledge management to

recommend user’s relevant products, services, or

contents (Park et al., 2012). Main data sources for

recommender systems are internal databases,

websites, social media, and integrated

devices/sensors (Fan, Lau, & Zhao, 2015; Park et al.,

2012). Data sources for RSs have changed along with

the prosperity of the era of information overload.

The first generation of RSs are built from website

data on purchases, users’ demography and preference

(Bobadilla et al., 2013). In the digital age, user data

can be initially cookies and server logs (Chen,

Chiang, & Storey, 2012). Enterprises can obtain

customer clickstream data logs on visit frequency,

viewed items, and visit time to demystify customer

behaviors (Fan et al., 2015; Park et al., 2012). Users’

demography and preference data reveal information

on age, gender, preferences, cultures, lifestyle,

purchasing power, shopping behaviors, customs and

habits of potential and existing customers (Albanese,

d'Acierno, et al., 2011; Dam, Le Dinh, & Menvielle,

2019). Therefore, the first generation of RSs will

allow enterprises to personalize recommendations for

each customer through data from clickstreams,

customer profiles, mobile call records, and

transactions for better customer satisfaction (Kabassi,

2013; Park et al., 2012).

The second generation of recommender systems

is built from social media data such as friends, likes,

followers, followed, tweets, and posts, etc. (Bobadilla

et al., 2013; Ma, Zhou, Liu, Lyu, & King, 2011).

Social intelligence extracted from user-contributed

data on social media can help enterprises in designing

new products, implementing marketing strategies,

and recommend relevant contents for online users

(Abbasi, Chen, & Salem, 2008; Archak, Ghose, &

Ipeirotis, 2011; Lau, Li, & Liao, 2014). The real-time

property of social intelligence along with its

subjectivity in a specific context is significantly

believed to be more trustworthy, updated, and reliable

compared to traditional sources of information

(Abbasi et al., 2008; Lau et al., 2014). Due to these

characteristics, data from social media can improve

the functions of prediction and recommendation of

recommender systems (Bobadilla et al., 2013; Ma et

al., 2011). Mining user-generated content and web

content will allow enterprises not only to develop

suitable products for customer needs but also to

predict a set of relevant products and recommend the

top ones to the right customers (Payne & Frow, 2005;

Rygielski, Wang, & Yen, 2002). Therefore, the

applications of recommender systems have emerged

in various fields, including music, books, shopping,

applications, TV programs, e-learning material, and

web search (Bobadilla et al., 2013; Park et al., 2012).

One of the challenges faced this generation is to

integrate and consolidate the different data sources

(Le Dinh & Nguyen-Ngoc, 2010).

The third generation of recommender systems is

built from data from integrated devices/sensors

(RIFD, real-time health, location-based, or weather

devices, etc.) as the trends of the Internet of Things

(Bobadilla et al., 2013). In particular, location-based

data which can be traced through mobile devices with

GPS, Wi-Fi, GSM, or Bluetooth, vehicles with GPS,

smart cards (bank cards or transportation cards),

floating sensors (devices with radio frequency

identification), check-in from social networks, can

offer real-time property for recommender systems

(Pan et al., 2013; Scellato, Noulas, & Mascolo, 2011).

These integrated devices /sensors, which represent

the Internet of Things, are catching attention in

building recommender systems (Bobadilla et al.,

2013; Ma et al., 2011).

5.3 Data Mining Models and

Algorithms

Different data mining models such as the association,

classification, clustering, and regression are

commonly used for building recommender systems

(Adomavicius & Tuzhilin, 2005). Association models

are widely used in many RSs with various data

ICEIS 2020 - 22nd International Conference on Enterprise Information Systems

722

mining techniques such as k-Nearest Neighbor,

Bayesian classifiers, association rules, decision trees,

link analysis, neural networks, and linear regression

(Adomavicius & Tuzhilin, 2005; Park et al., 2012).

As one of the most popular techniques, k-Nearest

Neighbor (k-NN) identifies users with similar

behaviors through their preference ratings and filters

top recommendations are most likely to be purchased

(Albadvi & Shahbazi, 2009). Other common data

mining techniques to develop RSs involve association

rules and link analysis (Park et al., 2012). On the other

hand, heuristics, clustering, and association models

are often applied to make recommendations related to

promotional strategies (Park et al., 2012).

Recommender systems will be able to promote

suitable products for target customers, especially for

movies and shopping industries (Payne & Frow,

2005; Rygielski et al., 2002). To sum up, the most

common data mining techniques to build

recommender systems are k-Nearest Neighbor,

association rules, and link analysis (Koren et al.,

2009; Park et al., 2012).

6 CONCLUSIONS

To facilitate the application of recommender systems

for small and medium-sized organizations and

enterprises (SMOs/SMEs) in the cultural domain, this

paper presents an in-depth literature review of 69

academic articles to gain an understanding of the

current status-quo and related aspects on data sources,

data mining models and algorithms. The concept-

centric approach along with the cluster analysis is

applied to categorize all the selected articles to

examine the coherence and relationships among

them. The results of this study have made important

theoretical and practical contributions.

The findings of this paper reveal the status of the

application of recommender systems in different

domains of the cultural sector. Despite the limited

number of articles in this research stream, articles

discussing the applications of RSs in each cultural

domain scatter all over the literature. Hardly any

paper has categorically synthesized and reviewed the

literature of all six cultural domains, including

heritage and libraries, live performance, visual and

applied arts, written and published works, audio-

visual and interactive media, and sound recording.

Therefore, this paper can be considered as the

pioneered literature review, which fulfills a

significant gap in the cultural sector. Another

theoretical contribution of this paper relates to the

application of information systems in the reflection of

recommender systems for cultural SMEs/SMOs.

Considering the fact that there is little research, which

links the two different domains - information systems

and the cultural sector, the proposed classification

framework can be a reference for researchers to

deepen their studies and enrich literature.

In terms of practical contributions, the literature

review will give cultural SMEs/SMOs an idea on

what type of recommender systems would suit their

needs. The paper also foresees the relevant issues

related to the data sources, algorithms, levels of user

engagement, and requirements for IT infrastructure.

The proposed future research directions also infer

enterprises about state-of-the-art RSs. Catching up

with the trend in the development of RSs will help

enterprises better respond to their needs. Therefore,

the paper will serve as a good starting point for

managers who consider adopting a recommender

system. The successful adoption of a recommender

system would enhance the discoverability of cultural

content and products for enterprises in the cultural

sector. Without any doubt that cultural SMEs/SMOs

which can take advantage of recommender systems

can gain a competitive advantage in the fierce

competition of the big data era.

At present, a conceptual framework for context-

aware recommender systems is being developed. This

framework will be experienced with a case study,

which aims at developing a recommender system for

a regional cultural development organization in

Canada to promote the discoverability of the products

and services of its members.

REFERENCES

Abbasi, A., Chen, H., & Salem, A. (2008). Sentiment

analysis in multiple languages: Feature selection for

opinion classification in Web forums. ACM

Transactions on Information Systems, 26(3), 12.

Adomavicius, G., & Tuzhilin, A. (2005). Toward the next

generation of recommender systems: A survey of the

state-of-the-art and possible extensions. IEEE

Transactions on Knowledge & Data Engineering(6).

Albadvi, A., & Shahbazi, M. (2009). A hybrid

recommendation technique based on product category

attributes. Expert Systems with Applications, 36(9),

11480-11488. doi:10.1016/j.eswa.2009.03.046

Albanese, M., d'Acierno, A., Moscato, V., Persia, F., &

Picariello, A. (2011). A multimedia semantic

recommender system for cultural heritage applications.

Paper presented at the 2011 IEEE Fifth International

Conference on Semantic Computing.

Albanese, M., d’Acierno, A., Moscato, V., Persia, F., &

Picariello, A. (2011). A novel strategy for

recommending multimedia objects and its application

A Literature Review of Recommender Systems for the Cultural Sector

723

in the cultural heritage domain. International Journal

of Multimedia Data Engineering Management Science,

2(4), 1-18.

Alharthi, H., Inkpen, D., & Szpakowicz, S. (2018). A

survey of book recommender systems. Journal of

Intelligent Information Systems, 51(1), 139-160.

Andjelkovic, I., Parra, D., & O’Donovan, J. (2019).

Moodplay: interactive music recommendation based on

artists’ mood similarity. International Journal of

Human-Computer Studies, 121, 142-159.

Archak, N., Ghose, A., & Ipeirotis, P. G. (2011). Deriving

the pricing power of product features by mining

consumer reviews. Management Science, 57(8).

Ardissono, L., Kuflik, T., & Petrelli, D. (2012).

Personalization in cultural heritage: the road travelled

and the one ahead. User modeling user-adapted

interaction, 22(1-2), 73-99.

Barragáns-Martínez, A. B., Costa-Montenegro, E.,

Burguillo, J. C., Rey-López, M., Mikic-Fonte, F. A., &

Peleteiro, A. (2010). A hybrid content-based and item-

based collaborative filtering approach to recommend

TV programs enhanced with singular value

decomposition. Information sciences, 180(22), 4290-

4311.

Bartolini, I., Moscato, V., Pensa, R. G., Penta, A.,

Picariello, A., Sansone, C., & Sapino, M. L. (2013).

Recommending multimedia objects in cultural heritage

applications. Paper presented at the International

Conference on Image Analysis and Processing.

Bartolini, I., Moscato, V., Pensa, R. G., Penta, A.,

Picariello, A., Sansone, C., & Sapino, M. L. (2016).

Recommending multimedia visiting paths in cultural

heritage applications. Multimedia tools applications,

75(7), 3813-3842.

Bauer, J., & Nanopoulos, A. (2014). Recommender systems

based on quantitative implicit customer feedback.

Decision Support Systems, 68, 77-88.

Bedi, P., & Agarwal, S. K. (2011). Managing security in

aspect oriented recommender system. Paper presented

at the 2011 International Conference on

Communication Systems and Network Technologies.

Bedi, P., & Vashisth, P. (2014). Empowering recommender

systems using trust and argumentation. Information

sciences, 279, 569-586.

Beel, J., Langer, S., Genzmehr, M., Gipp, B., Breitinger, C.,

& Nürnberger, A. (2013). Research paper

recommender system evaluation: a quantitative

literature survey. Paper presented at the Proceedings of

the International Workshop on Reproducibility and

Replication in Recommender Systems Evaluation.

Bobadilla, J., Ortega, F., Hernando, A., & Gutiérrez, A.

(2013). Recommender systems survey. Knowledge-

Based Systems, 46, 109-132.

doi:10.1016/j.knosys.2013.03.012

Borràs, J., Moreno, A., & Valls, A. (2014). Intelligent

tourism recommender systems: A survey. Expert

Systems with Applications, 41(16), 7370-7389.

Briguez, C. E., Budan, M. C., Deagustini, C. A.,

Maguitman, A. G., Capobianco, M., & Simari, G. R.

(2014). Argument-based mixed recommenders and

their application to movie suggestion. Expert Systems

with Applications, 41(14), 6467-6482.

Canada, S. (2011). Conceptual framework for culture

statistics 2011. In: Minister of Industry Ottawa, ON.

Carrer-Neto, W., Hernández-Alcaraz, M. L., Valencia-

García, R., & García-Sánchez, F. (2012). Social

knowledge-based recommender system. Application to

the movies domain. Expert Systems with Applications,

39(12), 10990-11000.

Champiri, Z. D., Shahamiri, S. R., & Salim, S. S. B. (2015).

A systematic review of scholar context-aware

recommender systems. Expert Systems with

Applications, 42(3), 1743-1758.

Chen, H., Chiang, R. H. L., & Storey, V. C. (2012).

Business intelligence and analytics: From big data to

big impact. MIS Quarterly, 36, 1165-1188.

Chianese, A., Marulli, F., Piccialli, F., & Valente, I. (2013).

A novel challenge into multimedia cultural heritage: an

integrated approach to support cultural information

enrichment. Paper presented at the 2013 International

Conference on Signal-Image Technology & Internet-

Based Systems.

Chianese, A., & Piccialli, F. (2016). A smart system to

manage the context evolution in the Cultural Heritage

domain. Computers Electrical Engineering, 55, 27-38.

Christensen, I. A., & Schiaffino, S. (2011). Entertainment

recommender systems for group of users. Expert

Systems with Applications, 38(11), 14127-14135.

Colombo-Mendoza, L. O., Valencia-García, R., Rodríguez-

González, A., Alor-Hernández, G., & Samper-Zapater,

J. J. (2015). RecomMetz: A context-aware knowledge-

based mobile recommender system for movie

showtimes. Expert Systems with Applications, 42(3),

1202-1222.

Cuomo, S., De Michele, P., Galletti, A., & Piccialli, F.

(2015). A cultural heritage case study of visitor

experiences shared on a social network. Paper

presented at the 2015 10th International Conference on

P2P, Parallel, Grid, Cloud and Internet Computing

(3PGCIC).

Dam, N. A. K., Le Dinh, T., & Menvielle, W. (2019).

Marketing Intelligence From Data Mining Perspective

– A Literature Review. International Journal of

Innovation, Management and Technology.

Dasgupta, A., Ghosh, A., Kumar, R., Olston, C., Pandey,

S., Tomkins, A., . . . Tomkins, A. (2007). The

discoverability of the web. Paper presented at the

Proceedings of the 16th international conference on

World Wide Web.

Deng, F., Ren, P., Qin, Z., Huang, G., & Qin, Z. (2018).

Leveraging Image Visual Features in Content-Based

Recommender System. Scientific Programming, 2018.

Deng, S., Wang, D., Li, X., & Xu, G. (2015). Exploring user

emotion in microblogs for music recommendation.

Expert Systems with Applications, 42(23), 9284-9293.

Eirinaki, M., Gao, J., Varlamis, I., & Tserpes, K. (2018).

Recommender systems for large-scale social networks:

A review of challenges and solutions. Future

Generation Computer Systems.

ICEIS 2020 - 22nd International Conference on Enterprise Information Systems

724

Ekstrand, M. D., Riedl, J. T., & Konstan, J. A. (2011).

Collaborative filtering recommender systems.

Foundations Trends in Human–Computer Interaction,

4(2), 81-173.

Fan, S., Lau, R. Y. K., & Zhao, J. L. (2015). Demystifying

Big Data Analytics for Business Intelligence Through

the Lens of Marketing Mix. Big Data Research, 2(1),

28-32. doi:10.1016/j.bdr.2015.02.006

Felício, C. Z., De Almeida, C. M., Alves, G., Pereira, F. S.,

Paixao, K. V., De Amo, S., & Barcelos, C. A. (2016).

Vp-rec: A hybrid image recommender using visual

perception network. Paper presented at the 2016 IEEE

28th International Conference on Tools with Artificial

Intelligence (ICTAI).

Garcia, I., Sebastia, L., & Onaindia, E. (2011). On the

design of individual and group recommender systems

for tourism. Expert Systems with Applications, 38(6),

7683-7692.

Hariadi, A. I., & Nurjanah, D. (2017). Hybrid attribute and

personality based recommender system for book

recommendation. Paper presented at the 2017

International Conference on Data and Software

Engineering (ICoDSE).

Horsburgh, B., Craw, S., & Massie, S. (2015). Learning

pseudo-tags to augment sparse tagging in hybrid music

recommender systems. Artificial Intelligence, 219, 25-

39.

Hu, Y., Xiong, F., Lu, D., Wang, X., Xiong, X., & Chen, H.

(2019). Movie collaborative filtering with multiplex

implicit feedbacks. Neurocomputing.

Hyung, Z., Lee, K., & Lee, K. (2014). Music

recommendation using text analysis on song requests to

radio stations. Expert Systems with Applications, 41(5),

2608-2618.

Jannach, D., Resnick, P., Tuzhilin, A., & Zanker, M.

(2016). Recommender systems—beyond matrix

completion. Communications of the ACM, 59(11), 94-

102.

Kabassi, K. (2013). Personalisation systems for cultural

tourism. In Multimedia services in intelligent

environments (pp. 101-111): Springer.

Kaminskas, M., & Ricci, F. (2012). Contextual music

information retrieval and recommendation: State of the

art and challenges. Computer Science Review, 6(2-3),

89-119.

Katarya, R., & Verma, O. P. (2017). An effective

collaborative movie recommender system with cuckoo

search. Egyptian Informatics Journal, 18(2), 105-112.

Kazienko, P., & Kolodziejski, P. (2006). Personalized

Integration of Recommendation Methods for E-

commerce. IJCSA, 3(3), 12-26.

Kim, H.-N., Alkhaldi, A., El Saddik, A., & Jo, G.-S. (2011).

Collaborative user modeling with user-generated tags

for social recommender systems. Expert Systems with

Applications, 38(7), 8488-8496.

Kim, H.-N., El-Saddik, A., & Jo, G.-S. (2011).

Collaborative error-reflected models for cold-start

recommender systems. Decision Support Systems,

51(3), 519-531.

Koren, Y., Bell, R., & Volinsky, C. (2009). Matrix

factorization techniques for recommender systems.

Computer(8), 30-37.

Lau, R. Y. K., Li, C., & Liao, S. S. Y. (2014). Social

analytics: Learning fuzzy product ontologies for aspect-

oriented sentiment analysis. Decision Support Systems,

65, 80-94. doi:10.1016/j.dss.2014.05.005

Le Dinh, T., & Nguyen-Ngoc, A. V. (2010). A conceptual

framework for designing service-oriented inter-

organizational information systems. Paper presented at

the Proceedings of the 2010 Symposium on Information

and Communication Technology.

Lee, W.-P., Chen, C.-T., Huang, J.-Y., & Liang, J.-Y.

(2017). A smartphone-based activity-aware system for

music streaming recommendation. Knowledge-Based

Systems, 131, 70-82.

Li, J., Xu, W., Wan, W., & Sun, J. (2018). Movie

recommendation based on bridging movie feature and

user interest. Journal of computational science, 26,

128-134.

Lin, Z. (2014). An empirical investigation of user and

system recommendations in e-commerce. Decision

Support Systems, 68, 111-124.

Liu, C.-L., & Chen, Y.-C. (2018). Background music

recommendation based on latent factors and moods.

Knowledge-Based Systems, 159, 158-170.

Lu, J., Wu, D., Mao, M., Wang, W., & Zhang, G. (2015).

Recommender system application developments.

Decision Support Systems, 74(C), 12-32.

doi:10.1016/j.dss.2015.03.008

Ma, H., Zhou, D., Liu, C., Lyu, M. R., & King, I. (2011).

Recommender systems with social regularization.

Paper presented at the Proceedings of the fourth ACM

international conference on Web search and data

mining.

Mathew, P., Kuriakose, B., & Hegde, V. (2016). Book

Recommendation System through content based and

collaborative filtering method. Paper presented at the

2016 International Conference on Data Mining and

Advanced Computing (SAPIENCE).

Moreno, A., Valls, A., Isern, D., Marin, L., & Borràs, J.

(2013). Sigtur/e-destination: ontology-based

personalized recommendation of tourism and leisure

activities. Engineering Applications of Artificial

Intelligence, 26(1), 633-651.

Moreno, M. N., Segrera, S., López, V. F., Muñoz, M. D., &

Sánchez, Á. L. (2016). Web mining based framework

for solving usual problems in recommender systems. A

case study for movies׳ recommendation.

Neurocomputing, 176, 72-80.

Nirwan, H., Verma, O. P., & Kanojia, A. (2016).

Personalized hybrid book recommender system using

neural network. Paper presented at the 2016 3rd

International Conference on Computing for Sustainable

Global Development (INDIACom).

Noguera, J. M., Barranco, M. J., Segura, R. J., & MartíNez,

L. (2012). A mobile 3D-GIS hybrid recommender

system for tourism. Information sciences, 215, 37-52.

Núñez-Valdéz, E. R., Lovelle, J. M. C., Martínez, O. S.,

García-Díaz, V., De Pablos, P. O., & Marín, C. E. M.

A Literature Review of Recommender Systems for the Cultural Sector

725

(2012). Implicit feedback techniques on recommender

systems applied to electronic books. Computers in

Human Behavior, 28(4), 1186-1193.

Pan, G., Qi, G., Zhang, W., Li, S., Wu, Z., & Yang, L. T.

(2013). Trace analysis and mining for smart cities:

issues, methods, and applications. IEEE

Communications Magazine, 51(6), 120-126.

Park, D. H., Kim, H. K., Choi, I. Y., & Kim, J. K. (2012).

A literature review and classification of recommender

systems research. Expert Systems with Applications,

39(11), 10059-10072. doi:10.1016/j.eswa.2012.02.038

Payne, A., & Frow, P. (2005). A strategic framework for

customer relationship management. Journal of

Marketing, 69(4), 167-176.

Pereira, A. L. V., & Hruschka, E. R. (2015). Simultaneous

co-clustering and learning to address the cold start

problem in recommender systems. Knowledge-Based

Systems, 82, 11-19.

Portugal, I., Alencar, P., & Cowan, D. (2018). The use of

machine learning algorithms in recommender systems:

A systematic review. Expert Systems with Applications,

97, 205-227.

Rickenberg, T. A., Neumann, M., Hohler, B., & Breitner,

M. (2012). Enterprise content management-A literature

review. Paper presented at the Proceedings of the

Eighteenth Americas Conference on Information

Systems.

Rygielski, C., Wang, J.-C., & Yen, D. C. (2002). Data

mining techniques for customer relationship

management. Technology in society, 24(4), 483-502.

Sanchez, F., Barrilero, M., Uribe, S., Alvarez, F., Tena, A.,

& Menendez, J. M. (2012). Social and content hybrid

image recommender system for mobile social networks.

Mobile Networks and Applications, 17(6), 782-795.

Sánchez-Moreno, D., González, A. B. G., Vicente, M. D.

M., Batista, V. F. L., & García, M. N. M. (2016). A

collaborative filtering method for music

recommendation using playing coefficients for artists

and users. Expert Systems with Applications, 66, 234-

244.

Sassi, I. B., Mellouli, S., & Yahia, S. B. (2017). Context-

aware recommender systems in mobile environment:

On the road of future research. Information Systems, 72,

27-61.

Scellato, S., Noulas, A., & Mascolo, C. (2011). Exploiting

place features in link prediction on location-based

social networks. Paper presented at the Proceedings of

the 17th ACM SIGKDD international conference on

Knowledge discovery and data mining.

Schafer, J. B., Frankowski, D., Herlocker, J., & Sen, S.

(2007). Collaborative filtering recommender systems.

In B. Peter, K. Alfred, & N. Wolfgang (Eds.), The

adaptive web (pp. 291-324): Springer-Verlag.

Sejal, D., Rashmi, V., Venugopal, K., Iyengar, S., &

Patnaik, L. (2016). Image recommendation based on

keyword relevance using absorbing Markov chain and

image features. International Journal of Multimedia

Information Retrieval, 5(3), 185-199.

Siu, N. Y.-M., Zhang, T. J.-F., Dong, P., & Kwan, H.-Y.

(2013). New service bonds and customer value in

customer relationship management: The case of

museum visitors. Tourism Management, 36, 293-303.

Tkalcic, M., Odic, A., Kosir, A., & Tasic, J. (2012).

Affective labeling in a content-based recommender

system for images. IEEE Transactions on Multimedia,

15(2), 391-400.

Umanets, A., Ferreira, A., & Leite, N. (2014). GuideMe–A

tourist guide with a recommender system and social

interaction. Procedia Technology, 17, 407-414.

van Capelleveen, G., Amrit, C., Yazan, D. M., & Zijm, H.

(2019). The recommender canvas: a model for

developing and documenting recommender system

design. Expert Systems with Applications, 129, 97-117.

Viktoratos, I., Tsadiras, A., & Bassiliades, N. (2018).

Combining community-based knowledge with

association rule mining to alleviate the cold start

problem in context-aware recommender systems.

Expert Systems with Applications, 101, 78-90.

Villegas, N. M., Sánchez, C., Díaz-Cely, J., & Tamura, G.

(2018). Characterizing context-aware recommender

systems: A systematic literature review. Knowledge-

Based Systems, 140, 173-200.

Webster, J., & Watson, R. T. (2002). Analyzing the past to

prepare for the future: Writing a literature review. MIS

Quarterly, xiii-xxiii.

Wedel, M., & Kannan, P. K. (2016). Marketing Analytics

for Data-Rich Environments. Journal of Marketing,

80(6), 97-121. doi:10.1509/jm.15.0413

Yang, W.-S., & Hwang, S.-Y. (2013). iTravel: A

recommender system in mobile peer-to-peer

environment. Journal of Systems Software, 86(1), 12-

20.

Yao, L., Sheng, Q. Z., Segev, A., & Yu, J. (2013).

Recommending web services via combining

collaborative filtering with content-based features.

Paper presented at the 2013 IEEE 20th International

Conference on Web Services.

Zhang, H.-R., & Min, F. (2016). Three-way recommender

systems based on random forests. Knowledge-Based

Systems, 91, 275-286.

Zhang, J., Yang, Y., Tian, Q., Zhuo, L., & Liu, X. (2017).

Personalized social image recommendation method

based on user-image-tag model. IEEE Transactions on

Multimedia, 19(11), 2439-2449.

Zheng, E., Kondo, G. Y., Zilora, S., & Yu, Q. (2018). Tag-

aware dynamic music recommendation. Expert Systems

with Applications, 106, 244-251.

Zhou, M. (2010). Book recommendation based on web

social network. Paper presented at the 2010

international conference on artificial intelligence and

education (ICAIE).

ICEIS 2020 - 22nd International Conference on Enterprise Information Systems

726