To search, Click below search items. |
| All Published Papers Search Service |
|
|
|
Title |
Addressing the Cold-start Problem Using Data Mining Techniques and Improving Recommendation By Cuckoo Algorithm: A Case Study of Facebook |
|
Author |
Saman Forouzandeh, Atae Rezaei Aghdam, Shuxiang Xu, and Elmira Akbari Pour Dibazar |
| Citation |
Vol. 26 No. 8 pp. 150-160 |
|
Abstract |
The popularity of Social networks, user demands, market realities, and technology developments are driving recommendation systems to explore new models of marketing and advertisements. Each user on social media sites shows a specific behavior such as; ranking interest items or like favorite contents. Hence, by studying user profiles we can explore their interests and preferences. Due to the great bulk of data on social media websites, the process of extracting hidden knowledge from data has become a hectic activity. For achieving this goal data mining techniques were flourished to discover interesting knowledge along with recommendation systems to suggest appropriate items to users based on this extracted knowledge. On of the most common obstacles in recommendation systems is cold-start problem, which is related to users who do not indicates any behavior on social medias. This paper aims to propose a solution for tackling this problem by using data mining techniques. Data were collected from facebook user accounts through Netvizz tools. In the next level, we enhance the recommendation method through Cuckoo algorithm to offer minimum number of items to get maximum feedback from users. Results indicate high performance of our propose solution. |
|
Keywords |
Recommendatiuon systems, Data mining, Social networks, Cuckoo algorithm, cold-start |
|
URL |
