Abstract
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.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
By “knowledge content” here we mean those contents that are conveying some piece of knowledge for the relevant users. Such contents are usually specific and specialized. The other characteristic of knowledge contents is that they can be retrieved and used at any time by relevant users and they do not usually expire over time.
- 2.
By “trendy content” here we mean those contents that are usually very general and can be relevant to the majority of users. The relevance of such contents, however, usually diminishes rapidly because they typically represent some events. Therefore, the interest of people in those contents decreases drastically over time.
- 3.
- 4.
- 5.
- 6.
- 7.
- 8.
- 9.
Name Entity Recognizers (NER) like Alchemy API, DBpedia Spotlight, Extractive, OpenClalais and Zamanta could be used for mapping extracted keywords to meaningful concepts or semantic labels. In this experiment we used DBpedia Spotlight.
- 10.
- 11.
- 12.
- 13.
- 14.
In our experiments we considered quantile of distribution of each concept (or label) similarities with all the other concepts (labels) in the collection as one of the metrics for setting threshold on the number of labels going to be augmented.
- 15.
(1) The Swiss A.I. Lab IDSIA (Istituto Dalle Molle di Studi sull’Intelligenza Artificiale), (2) Faculty of Informatics of University of Lugano, (3) Institute of Computational Science of the University of Lugano.
References
Kamran, S., Jazayeri, M.: Dynamic content and user identification in social semantic tagging systems. In: Luo, Y. (ed.) CDVE 2015. LNCS, vol. 9320, pp. 36–47. Springer, Cham (2015). doi:10.1007/978-3-319-24132-6_5
Braun, S., Schora, C., Zacharias, V.: Semantics to the bookmarks: a review of social semantic bookmarking systems. In: Proceedings of I-KNOW (2009)
Hotho, A., Jaschke, R., Schmitz, C., Stumme, G.: Bibsonomy: a social bookmark and publication sharing system. In: CS-TIW06. Aalborg University Press, Aalborg (2006)
Abel, F., Henza, F., Krause, D., Plappert, D., Siehndel, P.: GroupMe! where semantic web meets web 2.0. In: Proceeding of the 6th International Semantic Web Conference (2007)
Koivunen, M.: Semantic authoring by tagging with annotea social bookmarks and topics. In: Proceedings of the Semantic Authoring and Annotation Workshop at the International Semantic Web Conference, ISWC (2006)
Lachica, R., Karabeg, D.: Quality, relevance and importance in information retrieval with fuzzy semantic networks. In: Proceeding of TMRA 2008, University of Leipzig LIV (2008)
Zacharias, V., Braun, S.: SOBOLEO: social bookmarking and lightweight ontology engineering. In: Proceedings of the Workshop on Social and Collaborative Construction of Structured Knowledge at WWW 2007, vol. 273. CEUR Workshop Proceedings (2007)
Bollen, D., Halpin, H.: An experimental analysis of suggestions in collaborative tagging. In: IEEE/WICI/ACM International Joint Conference on Web Intelligence and Intelligent Agents Technology, Milan, Italy, pp. 108–115 (2009)
Yang, Y.H., Bogdanov, D., Herrera, P., Sordo, M.: Music retagging using label propagation and robust principal component analysis. In: Proceeding of the 21st International Conference on World Wide Web. ACM, New York (2012)
Kamran, S., Jazayeri, M., Ettefagh, T.: CODer entrepreneurs: how entrepreneurs can benefit from a dynamic social semantic tagging system. In: Proceedings of the 19th ACM Conference on Computer Supported Cooperative Work and Social Computing Companion, San Francisco, California, USA, pp. 305–308 (2016)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2017 Springer International Publishing AG
About this paper
Cite this paper
Kamran, S., Xu, Z., Jazayeri, M. (2017). Enhancing Accuracy of Dynamic Collaborative Labeling and Matching Through Semantic Augmentation Method (SAM). In: Luo, Y. (eds) Cooperative Design, Visualization, and Engineering. CDVE 2017. Lecture Notes in Computer Science(), vol 10451. Springer, Cham. https://doi.org/10.1007/978-3-319-66805-5_14
Download citation
DOI: https://doi.org/10.1007/978-3-319-66805-5_14
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-66804-8
Online ISBN: 978-3-319-66805-5
eBook Packages: Computer ScienceComputer Science (R0)