Social media are using several automatic, semi-automatic or even manual labeling approaches in order to match the shared contents with their users’ interests. The low degree of Click Through Rate (CTR) on the social media platform, however, suggests that labeling of shared contents and users’ active interests contain inaccuracies, leading to unsuccessful matching. One of the main reasons of unsuccessful matching is the heterogeneity of the labels assigned to the contents and users’ interest. In our previous work, we have proposed the Interactive and Dynamic Collaborative Labeling (IDCOLAB) framework in order to collect homogeneous and commonly agreed opinion of a group of users who are knowledgeable about the assigned labels dynamically. An essential step of IDCOLAB is Semantic Augmentation Method (SAM) which enables collaborative labeling of shared contents by dynamically augmenting semantically related labels to labels assigned initially to the contents and users’ interests. A goal of the augmentation process is to avoid irrelevant and noisy labels. We have applied SAM on COD which is a collaborative labeling platform based on IDCOLAB framework and evaluated SAM with two separate focus groups in the domains of Artificial Intelligence and Entrepreneurship.
CITATION STYLE
Kamran, S., Xu, Z., & Jazayeri, M. (2017). Enhancing Accuracy of Dynamic Collaborative Labeling and Matching Through Semantic Augmentation Method (SAM). In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 10451 LNCS, pp. 103–113). Springer Verlag. https://doi.org/10.1007/978-3-319-66805-5_14
Mendeley helps you to discover research relevant for your work.