A Personalization-Oriented Academic Literature Recommendation Method
As the number of digital academic items increases dramatically, it is more and more difficult for a student or researcher to find the expected references in a large academic literature database. Although collaborative filtering and content-based recommendation approaches perform well in some applica...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Ubiquity Press
2015-05-01
|
| Series: | Data Science Journal |
| Subjects: | |
| Online Access: | http://datascience.codata.org/articles/566 |
| _version_ | 1811193963047026688 |
|---|---|
| author | Zhongya Wang Ying Liu Jiajun Yang Zheng Zheng Kaichao Wu |
| author_facet | Zhongya Wang Ying Liu Jiajun Yang Zheng Zheng Kaichao Wu |
| author_sort | Zhongya Wang |
| collection | DOAJ |
| description | As the number of digital academic items increases dramatically, it is more and more difficult for a student or researcher to find the expected references in a large academic literature database. Although collaborative filtering and content-based recommendation approaches perform well in some applications, they do not produce satisfactory recommendations for academic items because they fail to reflect researchers’ unique characteristics in terms of authority, popularity, recentness, etc. In this paper, we propose two novel data structures, ALVector, which expresses various objective attributes of an article, and AUVector, which expresses users’ subjective weights for different attributes. Then, we propose a novel personalization-oriented recommendation method that utilizes both the content and non-content attributes in ALVector and AUVector for making recommendations. In order to make the overall best recommendation, the VIKOR algorithm is used with a personalization-oriented method to achieve a compromise solution. A real-world literature data set is used in the experiments. The experimental results show that our method better meets the user’s preference in multiple dimensions simultaneously. |
| first_indexed | 2024-04-12T00:18:15Z |
| format | Article |
| id | doaj.art-bfd8a64b346c4871a035a7b7a058a364 |
| institution | Directory Open Access Journal |
| issn | 1683-1470 |
| language | English |
| last_indexed | 2024-04-12T00:18:15Z |
| publishDate | 2015-05-01 |
| publisher | Ubiquity Press |
| record_format | Article |
| series | Data Science Journal |
| spelling | doaj.art-bfd8a64b346c4871a035a7b7a058a3642022-12-22T03:55:47ZengUbiquity PressData Science Journal1683-14702015-05-011410.5334/dsj-2015-017583A Personalization-Oriented Academic Literature Recommendation MethodZhongya Wang0Ying Liu1Jiajun Yang2Zheng Zheng3Kaichao Wu4School of Computer and Control, University of Chinese Academy of Sciences, BeijingSchool of Computer and Control, University of Chinese Academy of Sciences, Beijing Fictitious Economy and Data Science Research Center, Chinese Academy of Sciences, BeijingSchool of Computer and Control, University of Chinese Academy of Sciences, BeijingComputer and Network Information Center, Chinese Academy of Sciences, BeijingComputer and Network Information Center, Chinese Academy of Sciences, BeijingAs the number of digital academic items increases dramatically, it is more and more difficult for a student or researcher to find the expected references in a large academic literature database. Although collaborative filtering and content-based recommendation approaches perform well in some applications, they do not produce satisfactory recommendations for academic items because they fail to reflect researchers’ unique characteristics in terms of authority, popularity, recentness, etc. In this paper, we propose two novel data structures, ALVector, which expresses various objective attributes of an article, and AUVector, which expresses users’ subjective weights for different attributes. Then, we propose a novel personalization-oriented recommendation method that utilizes both the content and non-content attributes in ALVector and AUVector for making recommendations. In order to make the overall best recommendation, the VIKOR algorithm is used with a personalization-oriented method to achieve a compromise solution. A real-world literature data set is used in the experiments. The experimental results show that our method better meets the user’s preference in multiple dimensions simultaneously.http://datascience.codata.org/articles/566Recommendation systemPersonalizationOptimizationContent-based recommendation |
| spellingShingle | Zhongya Wang Ying Liu Jiajun Yang Zheng Zheng Kaichao Wu A Personalization-Oriented Academic Literature Recommendation Method Data Science Journal Recommendation system Personalization Optimization Content-based recommendation |
| title | A Personalization-Oriented Academic Literature Recommendation Method |
| title_full | A Personalization-Oriented Academic Literature Recommendation Method |
| title_fullStr | A Personalization-Oriented Academic Literature Recommendation Method |
| title_full_unstemmed | A Personalization-Oriented Academic Literature Recommendation Method |
| title_short | A Personalization-Oriented Academic Literature Recommendation Method |
| title_sort | personalization oriented academic literature recommendation method |
| topic | Recommendation system Personalization Optimization Content-based recommendation |
| url | http://datascience.codata.org/articles/566 |
| work_keys_str_mv | AT zhongyawang apersonalizationorientedacademicliteraturerecommendationmethod AT yingliu apersonalizationorientedacademicliteraturerecommendationmethod AT jiajunyang apersonalizationorientedacademicliteraturerecommendationmethod AT zhengzheng apersonalizationorientedacademicliteraturerecommendationmethod AT kaichaowu apersonalizationorientedacademicliteraturerecommendationmethod AT zhongyawang personalizationorientedacademicliteraturerecommendationmethod AT yingliu personalizationorientedacademicliteraturerecommendationmethod AT jiajunyang personalizationorientedacademicliteraturerecommendationmethod AT zhengzheng personalizationorientedacademicliteraturerecommendationmethod AT kaichaowu personalizationorientedacademicliteraturerecommendationmethod |