Encyclopedia of Computer Graphics and Games

Living Edition
| Editors: Newton Lee

3D Visualization Interface for Temporal Analysis of Social Media

  • Masahiko ItohEmail author
Living reference work entry
DOI: https://doi.org/10.1007/978-3-319-08234-9_42-1

Keywords

Social Medium Dependency Relation Tree Representation Textural Content Link Structure 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Definition

3D visualization interface for temporal analysis of social media is the interface for visual analytics of various types of time varying media contents using 3D information visualization techniques.

Introduction

Social media such as blogs and microblogs has become popular. It enables us to easily and rapidly publish information on our personal activities, interests, and opinions through writing document, creating links to other information resources, and providing images and/or movies. It dynamically reflects real movements in society.

Many organizations have collected and archived social media contents over the long term. Time series of archived data enable us to analyze temporal changes in trends in social media that reflect both real and virtual activities.

Visual analytics for extracting trends and reading stories from time sequential data sets are important research domains. There has been much research on analyzing temporal changes in trends on social media through visualizing link structures, results of text analysis, or flows of images (Kitsuregawa et al. 2008; Chi et al. 1998; Chi and Card 1999; Toyoda and Kitsuregawa 2005; Kehoe and Gee 2009).

In this entry, I will introduce three kinds of 3D information visualization systems for analyzing temporal changes in: (i) link structure, (ii) textural contents, and (iii) image contents on social media.

State-of-the-Art Work for Visualizing Temporal Changes in Social Media Contents

Visualization for Temporal Changes in Link Structure

We first introduce an interactive 3D visualization system for the time series of web graphs (Itoh et al. 2010). It is to enable us to examine the evolution of web graphs by comparing multiple graphs that have different timings and topics. To accomplish the system, it utilized interactive 3D components called TimeSlices that are 2D planes to visualize web graphs in a 3D environment. We can interactively add new TimeSlices along the timeline, and they can manipulate them to animate web graphs. Visualized web graphs on TimeSlices are snapshots of different timings.

Figure 1 shows the example for visualizing changes in link structure on blogs related to the term “working poor.” First peak appeared after the TV program called the “working poor” had been broadcast. Most blogs were linked to the official page of the TV program. The second peak appeared after the “working poor II” had been broadcast. We can find that influencers shifts in focus from the official pages of “working poor” to “working poor II.”
Fig. 1

Example for visualizing changes in link structure on blogs related to the term “working poor”

Visualization for Temporal Changes in Textural Contents

We next introduce an interactive 3D visualization system for exploring temporal changes in bloggers’ activities and interests through visualizing phrase dependency structures (Itoh et al. 2012). To accomplish the system, it utilizes two 3D components such as TimeSlices and TimeFluxes. TimeFluxes enable us to visualize temporal changes in the attribute values of particular nodes at every timing. The system visualizes dependency structures of phrases as a unified tree representation in TimeSlices and enables us to interactively navigate to the detailed information by expanding nodes in the tree representation. Sliding operation for the TimeSlices along the timeline indicates changes in the structure and frequencies of dependency relations. To compare different timings and topics side by side, it provides multiple 2D planes. It also visualizes changes in the frequencies of dependency relations by using TimeFluxes.

Figure 2 shows the example for visualizing changes in textural contents on blogs for comparing marketing effect of two telcos. The upper TimeSlice shows a topic for “Telco A,” while the lower one shows a topic for “Telco B.” (i) We can recognize events related to “change/switch to Telco A” are more popular than “change/switch to Telco B” in most months by observing changes in the structure and frequencies for events. (ii) We can expand nodes related to “announce” and “release” to find details on announcements and products that were released, and we then find that “Telco A” announced a “new price plan” in the first peak and released “product A” in the second peak.
Fig. 2

Example for visualizing changes in textural contents on blogs for comparing marketing effect of two telcos

Visualization for Temporal Changes in Image Contents

We finally introduce a 3D system for visualizing visual trends on social media that chronologically displays extracted clusters of images on blogs (Itoh et al. 2013). The system first adopts a histogram of images by stacking them on a timeline to visualize the flow of various images at each timing to visually analyze trends. This design enables us to find the timing for the beginning of the topic, changes in trends for the topic, bursting points, and a lifetime of the trends. Secondly, it arranges multiple histograms of images in a 3D space to visualize images on different aspects. This design allows us to observe different situations between different topics, sequences of trends, and events with the same timing on different topics.

Figure 3 shows the example for visualizing clusters of images related to “Prime Minister Hatoyama” extracted from blog based on visual, textual, and chronological similarities. The top 20 clusters are arranged from front to back according to their rankings. Images are aggregated per week. We can read stories about “Prime Minister Hatoyama” by exploring the movements of topics.
Fig. 3

Example for visualizing changes in clustered images related to “Prime Minister Hatoyama” extracted from blogs

Conclusion

This entry has introduced the 3D visualization systems for analyzing social media that utilized one dimension in a 3D space as a timeline. Although they independently visualized temporal changes in link structures, results of text analysis, and image clustering for single medium, we can combine these contents and mechanisms to construct integrated 3D visualization systems for intermedia analysis.

References

  1. Chi, E.H., Card, S.K.: Sensemaking of evolving web sites using visualization spreadsheets. In: Proceedings of InfoVis’99, pp. 18–25 (1999)Google Scholar
  2. Chi, E.H., Pitkow, J., Mackinlay, J., Pirolli, P., Gossweiler, R., Card, S.K.: Visualizing the evolution of web ecologies. In: Proceedings of CHI’98, pp. 400–407 (1998)Google Scholar
  3. Itoh, M., Toyoda, M., Kitsuregawa, M.: An interactive visualization framework for time-series of web graphs in a 3D environment. In: Proceedings of iV 2010, pp. 54–60 (2010)Google Scholar
  4. Itoh, M., Yoshinaga, N., Toyoda, M., Kitsuregawa, M.: Analysis and visualization of temporal changes in Bloggers’ activities and interests. In: Proceedings of PVis 2012, pp. 57–64 (2012)Google Scholar
  5. Itoh, M., Toyoda, M., Kitsuregawa, M.: Visualizing time-varying topics via images and texts for inter-media analysis. In: Proceedings of iV 2013, pp. 568–576 (2013)Google Scholar
  6. Kehoe, A., Gee, M.: Weaving web data into a diachronic corpus patchwork. Lang. Comput. 69(1), 255–279 (2009)Google Scholar
  7. Kitsuregawa, M., Tamura, T., Toyoda, M., Kaji, N.: Socio-sense: a system for analysing the societal behavior from long term web archive. In: APWeb; LNCS, vol. 4976, Springer, pp. 1–8 (2008)Google Scholar
  8. Toyoda, M., Kitsuregawa, M.: A system for visualizing and analyzing the evolution of the web with a time series of graphs. In: Proceedings of HYPERTEXT’05, pp. 151–160 (2005)Google Scholar

Copyright information

© Springer International Publishing Switzerland 2016

Authors and Affiliations

  1. 1.Institute of Industrial ScienceThe University of TokyoTokyoJapan