2013 Session: 791

• Activity Fragmentation, ICT, and Travel: Unraveling Interrelationships with Structural Equation Models
  Abstract: A substantial number of studies have addressed the relationship between Information and Communication Technologies (ICT), daily activities (here, paid work) and travel. Most studies have been primary concerned with direct effects of ICT on activities and travel. The aim of this study is to gain more insight into the relationship between ICT and travel behavior by using fragmentation as an intermediate concept to investigate how ICT influence travel behavior. The concept of fragmentation relates to how activities are reorganized temporally and spatially linked to ICT use. The causality of ICT, activity fragmentation and travel relationships remains to date unclear. We examine different causalities between ICT use, fragmentation and frequency of travel, based on a two-day communication-activity-travel data collected in The Netherlands. Using three different specifications, structural equation models (SEM) are applied to investigate the likely directions of the relationships. The results show that the causal associations between fragmentation ICT and travel are far from simple. ICT mediate the participation in non-work activities and can both substitute and complement the number of trips depending on the traveller’s attributes and type of ICT devices. More work fragmentation seems to limit the ability to travel for non-work purposes compared to work trips which is less elastic. ICT and fragmentation appear to have a reciprocal relationship with mobile ICT use influencing fragmentation while sedentary communications are more determined by the degree of fragmentation.
  Authors: Alexander, Bayarma; Ben-Elia, Eran; Ettema, Dick
  Authors: Alexander, Bayarma; Ben-Elia, Eran; Ettema, Dick
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Planning and Forecasting
  Session: 791
  Paper Number: 13-2057
• Influence of Social Contacts and Communication Use on Travel Behavior: Smart-Phone-Based Study
  Abstract: In this paper we investigate the use of a smartphone database to explore influences on travel behavior. Our aim is to exploit the rich individual-level data available from the smartphone to study the influence of communication and social contacts (collected via phone call and sms logs) on spatial movement (collected via GPS). An advantage of smartphone data is the ability to collect such rich data without user input over a long period of time, and the disadvantage is the difficulty associated with processing the data. We work with three months of data from 111 people collected via a snowball sample. In studying travel behavior, we focus on high level measures of mobility as represented by the size of one’s activity space and one’s travel intensity (our dependent variables). We use as explanatory variables sociodemographics, spatial relationship between home and work, communication use (number of phone calls and sms), and the travel behavior of those in the sample who are connected to the respondent (where connectivity is measured by phone and sms contact). We describe how these variables were processed from the smartphone data and present estimation results from the regression analysis. We find that people tend to travel in a similar manner as those they are socially connected to (consistent with the social network and travel literature) and that communication use is a compliment to physical travel (consistent with the telecommunication and travel literature). The results, although preliminary, illustrate how smartphone data can be exploited to reveal complex features of travel behavior.
  Authors: Ythier, Jeanne; Walker, Joan L.; Bierlaire, Michel
  Authors: Ythier, Jeanne; Walker, Joan L.; Bierlaire, Michel
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Planning and Forecasting
  Session: 791
  Paper Number: 13-4464
• Semantic Annotation of GPS Traces: Activity Type Inference
  Abstract: Due to the rapid development of technology, increasingly larger travel and activity behavior data exists to date. These large data sets often lack semantic interpretation, implying that annotation in terms of the activity being performed or the transportation mode being used is necessary. This paper aims to infer activity types from GPS traces by developing a decision tree-based model that considers activity start times and activity durations. Two models, i.e. a predicted probability distribution and a point prediction model, were derived from a decision tree classification. Two types of data were used, namely paper-and-pencil activity-travel diary data and corresponding GPS data. The data were collected in 2006 and 2007 in Flanders, Belgium. The most optimal classification tree constructed when considering both in-home and out-of-home activities comprises 18 leaves. Consequently, 18 if-then rules were derived. An accuracy of 74% was achieved when training the tree. The accuracy of the model for the validation set, i.e. 72.5%, shows that overfitting is minimal. When applying the model to the test set, the performance was almost 76% accurate. Based on the decision tree, a probability matrix was constructed. From this probability matrix, a point prediction was extracted using the highest probabilities per class. The models constructed indicate the importance of time information in the semantic enrichment process. The contribution of this study towards future data collection is promising in that it enables researchers to automatically infer activities solely from activity start time and duration information obtained from GPS data.
  Authors: Reumers, Sofie; Liu, Feng; Janssens, Davy; Cools, Mario; Wets, Geert
  Authors: Reumers, Sofie; Liu, Feng; Janssens, Davy; Cools, Mario; Wets, Geert
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Planning and Forecasting
  Session: 791
  Paper Number: 13-4496
• Using Smart Phones and Sensor Technologies to Automate Collection of Travel Data
  Abstract: This paper presents a comprehensive framework and its prototype application for activity-travel data collection through the exploitations of smartphone sensors. The core components of the framework run on smartphones, backed by cloud-based services for data storage, information dissemination and online decision support. The framework employs machine learning techniques to automatically infer activity types and travel modes with minimum interruptions for the respondents. There are three main components of the framework: 1) a 24 hours location data collection, 2) a dynamic land-use database, and 3) a transportation mode identification component. The location logger is based on the smartphone network and it can run for 24 hours with minimum impact on smartphone battery and equally applicable in places where GPS is available or not available. The land-use information is continuously updated from internet location services such as Foursquare. Transportation mode identification module is able to distinguish six different modes with 98.85% accuracy. Prototype application is conducted in the city of Toronto and results clearly indicate the viability of this framework.
  Authors: Abdulazim, Tamer; Abdelgawad, Hossam; Nurul Habib, Khandker M.; Abdulhai, Baher
  Authors: Abdulazim, Tamer; Abdelgawad, Hossam; Nurul Habib, Khandker M.; Abdulhai, Baher
  Year: 2013
  Document Type: Paper
  Subject: Data and Information Technology; Planning and Forecasting
  Session: 791
  Paper Number: 13-5254
• Activity Fragmentation, ICT, and Travel: Unraveling Interrelationships with Structural Equation Models
  Authors: Ben-Elia, Eran
  Authors: Ben-Elia, Eran
  Year: 2013
  Document Type: Presentation
  Subject: Data and Information Technology; Planning and Forecasting
  Session: 791
  Paper Number: 13-2057
• Influence of Social Contacts and Communication Use on Travel Behavior: Smart-Phone-Based Study
  Authors: Walker, Joan
  Authors: Walker, Joan
  Year: 2013
  Document Type: Presentation
  Subject: Data and Information Technology; Planning and Forecasting
  Session: 791
  Paper Number: 13-4464
• Semantic Annotation of GPS Traces: Activity Type Inference
  Authors: Reumers, Sofie
  Authors: Reumers, Sofie
  Year: 2013
  Document Type: Presentation
  Subject: Data and Information Technology; Planning and Forecasting
  Session: 791
  Paper Number: 13-4496
• Using Smart Phones and Sensor Technologies to Automate Collection of Travel Data
  Authors: Abdulazim, Tamer
  Authors: Abdulazim, Tamer
  Year: 2013
  Document Type: Presentation
  Subject: Data and Information Technology; Planning and Forecasting
  Session: 791
  Paper Number: 13-5254