On a Serendipity-oriented Recommender System based on Folksonomy and its Evaluation

Yamaba, Hisaaki; Tanoue, Michihito; Takatsuka, Kayoko; Okazaki, Naonobu; Tomita, Shigeyuki

doi:10.1016/j.procs.2013.09.104

Cited by 15 publications

(5 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We do not use book features but focusing on book author's features to improve serendipitous recommendations. Yamaba et al [5] proposed a serendipity-oriented recommender system based on Folksonomy. Yamaba used Folksonomy as an indicator to recommend serendipitous items.…”

Section: Serendipity In Rssmentioning

confidence: 99%

Book Recommender System Using Linked Data for Improving Serendipity

Weng

Takaku

2019

Journal of Japan Society of Information and Knowledge

View full text Add to dashboard Cite

Recent years, to overcome the flood of information, recommender systems (RSs) are being used in many scenarios, such as online shopping stores, movie website and so on. However, many recommender algorithms focus on accuracy based on a user profile, which may lead to reducing user's satisfaction. As high-accuracy based RSs suggest similar items that the user may have known before. As a result, the recommendation leads to hurt user's satisfaction. And there is a concept called serendipity which is a way to address this problem. This paper focuses on improving the serendipity of the RSs. We use two approaches, the variety of resources in Linked Data and author similarity. We extract author information from DBpedia dataset, which is one of the datasets in Linked Data. Then we calculate author similarity based on this information. This paper describes methods of book recommendations from the Book-Crossing dataset, reports recommendation results and discusses our recommender system according to our results.

show abstract

Section: Serendipity In Rssmentioning

confidence: 99%

Book Recommender System Using Linked Data for Improving Serendipity

Weng

Takaku

2019

Journal of Japan Society of Information and Knowledge

View full text Add to dashboard Cite

show abstract

“…It also includes preferences for 11 types of movies, 22 personal attribute variables, and 5 demographic variables. Using these observations, we will display the relationships between movie preferences and demographic information including age, gender and other factors that can be used to accurately recommend specific genres to customers as other recommendation systems did [18][19][20][21]. It has been approved that user information such as gender, location, or preference is effective to recommend movies to the customer [3,[22][23][24].…”

Section: Handle Datamentioning

confidence: 99%

Utilizing Imbalanced Data and Classification Cost Matrix to Predict Movie Preferences

Wang¹

2018

IJAIA

View full text Add to dashboard Cite

In this paper, we propose a movie genre recommendation system based on imbalanced survey data and unequal classification costs for small and medium-sized enterprises (SMEs) who need a data-based and analytical approach to stock favored movies and target marketing to young people. The dataset maintains a detailed personal profile as predictors including demographic, behavioral and preferences information for each user as well as imbalanced genre preferences. These predictors do not include movies' information such as actors or directors. The paper applies Gentle boost, Adaboost and Bagged tree ensembles as well as SVM machine learning algorithms to learn classification from one thousand observations and predict movie genre preferences with adjusted classification costs. The proposed recommendation system also selects important predictors to avoid overfitting and to shorten training time. This paper compares the test error among the above-mentioned algorithms that are used to recommend different movie genres. The prediction power is also indicated in a comparison of precision and recall with other state-of-the-art recommendation systems. The proposed movie genre recommendation system solves problems such as small dataset, imbalanced response, and unequal classification costs.

show abstract

“…The conventional classifications are top down approach but the folksonomy is the bottom up classification technique [7]. This technique also uses the tag information on the items assigned by the users.…”

Section: Related Work On Serendipity Folksonomy Basedmentioning

confidence: 99%

“…To fulfil it if we plot the curve of FP to FN then it gives information of not preferred but recommended items. This curve gives information about unexpected recommended items which leads to serendipity [7].…”

Section: Precision Recall F1 and Rocmentioning

confidence: 99%

Serendipity in Recommender Systems

Patel¹,

Amin²

2018

IJET

View full text Add to dashboard Cite

Today the growth of information technology, social networking and commercial websites are heavily increased. Because of it the information available for users is increased. Because of this information overload users cannot select best items from available various items. To solve this problem Recommender System (RS) required to explicit preferences and monitoring of implicit behaviour of user. The traditional RS recommends the items to the user based on their choice. But after some time user become bored with same kind of recommendations again and again. To provide surprise in the presented items another king of qualitative parameter required namely-"Serendipity". Serendipity refers to the "phenomenon of spontaneously understanding unexpected things, including time, location, people, and contents". This paper is detailed survey about serendipity in various aspects in study, equations, components, related works and measures of serendipity in RS. Keyword-Serendipity, Recommender System, Information Overload, Item Preferences I. INTRODUCTION The Recommender Systems recommends items, which might be useful for users. However, these items are not interesting if user is already familiar with the same product. Therefore, the novelty and serendipity in the recommended list are essential to make the recommendation interesting [12]. In current situation because of explosion of items in social networks and ecommerce, the research in serendipitous recommendation is essential. The word "Serendipity" comes from Old Persian tale whose name is "The Three Princes of Serendip". This tale was translated and publish by Cristoforo Armeno in 1557. Then after almost two hundred years, Horace Walpole use the word "Serendipity" means that to make the discovery by accident. The use of this terminology in the field of science by Foster and Ford in 2003. However, in RS research, serendipity means to recommend unexpected, unforeseeable or surprising items [12]. The recommended items to the user are serendipitous if items are based on user's test, are unknown and they are difficult to search it. Traditional RS only focus on the accuracy for recommending items. Recent studies suggests that beyond the accuracy other factors also should be considered for recommendation. Accuracy is one of the most required criteria but to present variety in the recommended list is also important concern [22]. Researcher says that instead of presenting the list of items in the recommendation with same properties and contents to list the items with different test and properties will more satisfies the users [17]. Before submitting your final paper, check that the format conforms to this template. Specifically, check the appearance of the title and author block, the appearance of section headings, document margins, column width, column spacing and other features.

show abstract

On a Serendipity-oriented Recommender System based on Folksonomy and its Evaluation

Cited by 15 publications

References 7 publications

Book Recommender System Using Linked Data for Improving Serendipity

Book Recommender System Using Linked Data for Improving Serendipity

Utilizing Imbalanced Data and Classification Cost Matrix to Predict Movie Preferences

Serendipity in Recommender Systems

Contact Info

Product

Resources

About