Using recommendations in content curation

Zurina , Saaya (2014) Using recommendations in content curation. PhD thesis, UTeM.

[img] Text
Using Recommendations In Content Curation 24pages.pdf

Download (5MB)


The evolution of information sharing on the web has introduced a new chapter for information discovery. With all of the information and all of the people together in one place, there are more opportunities for creating, sharing, and discovering information. Now, many individuals are using their Twitter and Facebook accounts to share interesting pieces of content they locate. To some extent this is known as content curation, which involves users as curators who search, �lter, organise and share the information they �nd. There is also a speci�c sites for content curation (e.g. Storify, Pinterest,, BagTheWeb) which provide users with a set of tools to manually collect, manage topical collections of content and share the content with others. As it stands though curation is very much a manual process with the user solely responsible for performing each of the aforementioned steps in curating collections of content. We believe that we can alleviate some of this burden on the user by providing intelligent assistance at di�erent stages of the curation cycle. In particular we focus on the search and organisation stages and identify two key tasks, assignment and discovery. The assignment task involves situating new content within a collection of other related content. In this thesis we endeavour to automate this process and identify the correct collection for incoming content as it is discovered by the user, thus making the process both simpler and more e�cient. We investigate recommender systems approaches and evaluate their e�cacy for two di�erent types of curation systems. The �rst,, is a traditional online curation service where users can both curate their own collections of content, and follow the collections of others. The second, HeyStaks, is a social search platform in which curation is directly integrated within everyday search. In HeyStaks communities of like-minded searchers can share curated repositories of search experiences. The discovery task involves identifying new and interesting content for a user to curate. We examine this task within the context of In particular, we exploit the information of collections that users have both curated and followed in order to establish their interests and recommend new collections for them to follow. By improving the manner in which content is organised and discovered we believe this research will help existing curators, encourage new curators, and improve the quality of content collections in general. An increase in both the quantity and quality of curated collections should in turn bene�t information seekers for whom search is increasingly not stringent enough in terms of seeking out the very best information

Item Type: Thesis (PhD)
Uncontrolled Keywords: Peer-to-peer architecture (Computer networks), Wireless communication systems
Subjects: T Technology > TK Electrical engineering. Electronics Nuclear engineering
Divisions: Library > Tesis > FTMK
Depositing User: Norziyana Hanipah
Date Deposited: 04 Sep 2015 07:41
Last Modified: 04 Sep 2015 07:41
Statistic Details: View Download Statistic

Actions (login required)

View Item View Item