Placegram: A Diagrammatic Map for Personal Geotagged Data Browsing Hyungeun Jo, Student Member, IEEE, and Jung-hee Ryu, Member, IEEE Abstract—Geotagging personal data such as photos and videos are continuously becoming easier and more popular. Nevertheless, browsing such data on general purpose maps can be difficult, due to the frequent zoom and pan operations as well as visual components unnecessary for the task. This paper presents Placegram, a compact diagrammatic map visualization for personal geotagged data browsing based on cognitive map theories. An evaluation using real-life data sets shows that the speed of finding and pointing to places from the participants’ own data increased by a factor of 2.1-2.9, and the number of interesting places discovered from others’ data within a time limit increased by 48.8 percent in Placegram compared to a general purpose map. Placegram was even slightly faster than a simple text list, while at the same time, preserving the geographic senses of direction and location. Subjective ratings and comments from participants support these results, indicating that Placegram is significantly preferred over both a general map and a text list. Index Terms—Information visualization, graphical user interfaces, spatial databases and GIS, multimedia information systems. Ç 1 INTRODUCTION G EOTAGGING is becoming an increasingly popular method for indexing and retrieving personal data such as photos and videos. In a number of photo sharing services or organization tools such as Flickr and Picasa, one can attach to a photo a geotag, a reference to which it was taken, and later retrieve the photo by location. The importance of geotagging is supported by autobiographical memory studies [35], [40], showing that where an event occurred is one of most well- remembered traits of the personal event, as well as an effective cue to elicit other traits such as what, who, and when. Most geotagging processes can be automated with current location-aware devices, and the importance of geotagging is expected to grow with the prevalence of such devices. Personal geotagged data are usually browsed on Web- based, general purpose maps. However, browsing such data in general purpose maps leads to several usability problems. First, frequent zoom and pan operations are required for even a small amount of data, because personally important places tend to be unevenly distributed on a map, as shown in Fig. 2a. The user must closely zoom in to browse a dense area, then zoom out, pan, and zoom in again to browse another dense area. Second, general purpose maps contain as detailed information about every place as possible. However, many of these details have no direct relation to the specific places the user is interested in, and hinder quick perception of the information such as the names of the places of interest and the names of the areas where these places are located. In response to these problems, this paper presents Placegram, a compact diagrammatic map visualization for personal geotagged data browsing, motivated from cogni- tive map theories. Cognitive maps refer to “mental representations of maps or environments” [38], and have received attention from various fields since Lynch [19] showed that they can largely differ from the actual structure of the area. Fig. 1 humorously illustrates the concept of the cognitive map. The distortions in cognitive maps are systematically induced both while memorizing the spatial information and processing it to mentally construct the representation [38]. These include, for example, scales according to a person’s perspective, places aligned horizontally and vertically, and directions in multi- ples of 90 or 45 degrees (reviewed in [39]). Placegram takes advantage of the concept of cognitive maps by adopting such mental distortions and visualizing them in a schematized way. By this strategy, Placegram achieves a visual cleanness and compact layout that dramatically minimizes the need to zoom and pan (Fig. 2b). This strategy also enables easier browsing in various mobile devices by providing natural mapping with four-directional keys, as well as a compact layout that fits in small screens (Fig. 2c). In the rest of this paper, we describe related work, design considerations, and the process of constructing Placegram. We then present the evaluation process and results that compare Placegram to a general purpose map and a simple text list. 2 RELATED WORK 2.1 Geotagging and Map-Based Browsing Research on geotagging personal data and browsing such data by location have been mainly focused on photos. Survey results by Naaman et al. showed that locations and people are the most well-remembered cues in recalling photos and are useful cues for finding photos, rather than temporal information such as year, month, and day [24]. IEEE TRANSACTIONS ON VISUALIZATION AND COMPUTER GRAPHICS, VOL. 16, NO. 2, MARCH/APRIL 2010 221 . The authors are with the Graduate School of Culture Technology, Korea Advanced Institute of Science and Technology, 335 Gwakak-ro, Yuseong-gu, Daejeon 305-701, Republic of Korea. E-mail: {acid, junghee.ryu}@kaist.ac.kr. Manuscript received 29 Aug. 2008; revised 18 Feb. 2009; accepted 1 June 2009; published online 10 June 2009. Recommended for acceptance by M.Q. Ward. For information on obtaining reprints of this article, please send e-mail to: tvcg@computer.org, and reference IEEECS Log Number TVCG-2008-08-0135. Digital Object Identifier no. 10.1109/TVCG.2009.68. 1077-2626/10/$26.00 ß 2010 IEEE Published by the IEEE Computer Society Authorized licensed use limited to: Korea Advanced Institute of Science and Technology. Downloaded on April 21,2010 at 02:28:50 UTC from IEEE Xplore. Restrictions apply.