Kelkar et al., ICWSM2007

Citation

Kelkar, S., A. John, and D. Seligmann. 2007. An Activity-based Perspective of Collaborative Tagging. In International Conference on Weblogs and Social Media, Boulder, Colorado.

[[1]]

Abstract from paper

Collaborative tagging offers an interesting framework for studying online activity as users, topics (tags), and resources (bookmarks) get associated with each other through a folksonomy. In this paper, we consider an activity-based perspective of collaborative tagging where activity is defined as the act of associating a tag with a bookmark by a user. The perspective categorizes activities based on two defined measures: intensity and spread, which indicate the level and range, respectively, of the tagging activity, measured for both users and tags. Our block-model perspective juxtaposes two subperspectives: (i) A user perspective that captures the activity of users across different tags and, (ii) A tag perspective that captures the activity in tags across different users. This juxtaposition can provide an insight into different communities of users and tags. It has applications in identifying trends and types of interests in web communities as well as expertise, staffing needs and knowledge gaps in enterprise communities. Results obtained by analyzing data from a commercial tagging service offer interesting case studies.

Summary

The authors developed an activity model for collaborative tagging.

Concepts and properties:

• Activity: User associate a tag with a bookmark to user
• Nature of bookmarks: 3 layers (users -> tags -> content)
• Connection within a layer: Two entities in a layer are never directly connected
• Connection between layers: user and content cannot be directly linked
• Nature of connections: Connection between layers are potentially unconstrained.
• Position: a set of individuals who are similarly embedded in network of relations with other individuals.
• Role: relationship with other actors or positions.

Their model:

• Try to reduce sets of actors to finite number of positions based on their activities

(I=intensity, S=spread, H=high, M=medium, L=low)

• • 4 for users: HI+HS, HI+LS, MI, LI
  • 4 for tags: HI+HS, HI+LS, MI, LI

Algorithm for position classification is explicitly defined using rule-like formulas in the paper. Dataset: RawSugar www.rawsugar.com social bookmarking site similar to Delicious.

The results, in my opinion, is unclear. It shows the trends and migration between categories( They even use two words "position" and "category" to describe one concept (or two similar concepts)). I do not think the authors made any solid or useful conclusions. It is more likely to be a showcase of distribution trends of the categories they defined in hands.

Related Work

• M. Dubinko, R. Kumar, J. Magnani, et al. (2006). Visualizing Tags over Time, Proceedings of the WWW, Edinburgh, Scotland, May 23-26
• S. Golder and B. Huberman (2005). The Structure of Collaborative Tagging Systems. HP Labs, 2005.

Study Plan

• Understand the data nature of RawSugar [[2]]
• Read Golder et al.'s paper, and understand the structure of the user-tagging network. [[3]]