TrailView: Combining Gamification and Social Network Voting Mechanisms for Useful Data Collection

There is a dearth of structured, organized photographical information of hiking trails and nature in general. Although people frequently photograph these locations and some efforts have been conducted to create virtual walk-throughs of select locations using specialized equipment (Stranger 2013; Olanoff 2013), the information is largely scattered across different social networks and other websites. With TrailView we aim to utilize the efforts and photographs of everyday hikers to create a structured view of hiking trails and nature. TrailView boasts a gamified system to encourage user-interaction and focuses on social interactions, feedback, and competition to drive data collection. TrailView also contains a point/incentive scheme designed to motivate users. This incentive scheme relies on a social voting model to drive useful data collection, and global leaderboards to encourage competition. TrailView: A Game With a Purpose Games with a purpose center around the construction of games where users, as a side effect of playing, perform tasks computers are unable to perform. Games with a purpose have, in recent years, been created to crowdsource the tagging and indexing of images (von Ahn and Dabbish 2004), and aid computers in folding proteins (Cooper et al. 2010). The application presented in this paper, TrailView, attempts to crowdsource and gamify the acquisition of photos for hiking trails and, in doing so, acquire trail information. The gamified system described in this paper aims to drive user engagement, content generation, and structured data. TrailView is a system designed to provide structured photographical information of hiking trails, mountains, and nature. The goal is to acquire enough information to create virtual walk-throughs of hiking trails throughout the world similar to what Google has done for streets with Street View. Unlike Google’s Street View, the acquisition of information cannot be automated by using motorized vehicles along the trails as this would negatively impact the quality of the hiking trails. Thus, we argue that hiking photos require a distributed, crowdsourced acquisition method. Copyright c © 2013, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved. The contributors to TrailView will be hikers and adventurers of all levels. However, the target users of TrailView can range from hikers looking to plan their next hike, or casual users looking to browse through some beautiful hiking trails. Much as Google Maps users use the service to plan a route, TrailView can be used to plan a hiking trail and preview the terrain that lays ahead. TrailView can also be used as a storage engine to keep track of personal photos from hikes and place them geographically and chronologically on a map. TrailView will utilize mobile applications and a website to reach the target contributors and users. It is designed to be minimally invasive, creating a thin layer between normal hiking routines and the TrailView data collection layer so as to not interfere with existing hiking patterns. Figure 1 below shows the typical TrailView user cycle. Users examine the GPS map to view potential trails. Trail photos are examined and used to both decide on a trail and see what areas have been photographed. The user then goes on a hike and takes new photos which are uploaded to TrailView. The corresponding photos earn points, badges, and titles. The user then returns to the map to begin the planning cycle for subsequent hikes. Figure 1: The TrailView user cycle. Social, Competition and Feedback As detailed by Huang and Fu (Huang and Fu 2012) and PhotoCity (Tuite et al. 2011), social aspects, competition, 80 Human Computation and Crowdsourcing: Works in Progress and Demonstration Abstracts AAAI Technical Report CR-13-01