Hung-Jen Yang

Contemporary Issues in Handheld Computing Research

Mobile phones have become ubiquitous in todays society. However, mobile users are no longer satisfied with simple phones but instead expect ever more powerful functions to be available from their mobile devices. Advanced phones known as smartphones allow mobile users to perform a wide variety of advanced handheld functions such as browsing the mobile Internet or finding a nearby theater showing a specific movie. The design and development of these new, improved handheld functions require the help of handheld computing research. This article introduces handheld computing research using three themes: i mobile handheld devices, ii mobile computing, and iii current issues in handheld computing research. Information about other handheld topics of interest to researchers is given in the last section.

Secure spatial trajectory prediction based on traffic flows

The ubiquity of smartphones has many developers join the mobile app development. Location-based services (LBSs), one of the mobile applications, have attracted a great attention recently. This research proposes a location-based service, spatial trajectory prediction, which is to predict the forthcoming locations of the human travel routes. It could be used in a variety of purposes such as travel recommendations and traffic control and planning. The proposed method is based on the previous traffic flows, a weighted graph. It predicts the spatial trajectory by finding a path with a maximal weight and within a length. A modified algorithm of the longest path is used to find the maximum-weight path. If the length of the path is over the limit specified by the user, the next maximum-weight path is used and the length is checked again. The process is repeated if necessary. Experimental results show the proposed method works, but further tests are needed to validate its effectiveness.

Privacy-Preserving Spatial Trajectory Prediction Based on a Novel Matrix Representation

Since the introduction of iPhone in 2007, smartphones have become very popular (e.g., the number of worldwide smartphone sales has surpassed the number of PC sales in 2011). The feature of high mobility and small size of smartphones has created many applications that are not possible or inconvenient for PCs and servers, even laptops. Location-based services (LBS), one of mobile applications, have attracted a great attention recently. This research proposes a location-based service, which predicts a spatial trajectory based on the current and previous trajectories by using a novel matrix representation. Spatial trajectory prediction can be used in a variety of purposes such as travel recommendations and traffic control and planning, but at the same time, just like most location-based services, the user privacy concern is a major issue. Without rigorous privacy protection, users would be reluctant to use the service. The proposed method is simple but effective and user privacy is rigorously preserved at the same time because the trajectory prediction is performed at the user-side. Additionally, this research is not only useful but also pedagogical because it involves a variety of topics like (i) mobile computing, (ii) mobile security, and (iii) human behavior recognition.

Incremental location searching for route anomaly detection

Since the introduction of iPhone in 2007, smartphones have become very popular. The high mobility of smartphones has created many applications that are not possible or inconvenient for PCs and servers, even notebooks. One of the mobile applications, location-based services (LBS), has attracted great attention recently. This paper proposes location-based research, which uses location information to find route anomalies, a common problem of daily life. For example, an alert should be generated when a school bus misses part of a route. Different kinds of route anomalies are discussed and various methods for detecting the anomalies are proposed in this paper. The major methods use a technique of incremental location search, which finds matched routes as the search route is entered location by location. An alert is generated when no matched routes exist. Preliminary experiment results show the proposed methods are effective and easy-to-use.

A novel matrix representation for privacy-preserving spatial trajectory prediction

Location-based services (LBS), one of mobile applications, have attracted a great attention recently. This research proposes a location-based service, which predicts a spatial trajectory based on the current and previous trajectories by using a novel matrix representation. Spatial trajectory prediction can be used in a variety of purposes such as travel recommendations and traffic control and planning, but at the same time, just like most location-based services, the user privacy concern is a major issue. Without rigorous privacy protection, users would be reluctant to use the service. The proposed method is simple but effective and user privacy is rigorously preserved at the same time because the trajectory prediction is performed at the user-side. Additionally, this research is not only useful but also pedagogical because it involves a variety of topics like (i) mobile computing, (ii) mobile security, and (iii) human behavior recognition.

Route Anomaly Detection Using a Linear Route Representation

A location-based service is a service based on the geographical position of a mobile handheld device like a smartphone. This research proposes location-based research, which uses location information to find route anomalies, a common problem of daily life. For example, an alert should be generated when a deliveryman does not follow his regular route to make deliveries. Different kinds of route anomalies are discussed and various methods for detecting the anomalies are proposed in this paper. The proposed method based on a linear route representation finds the matched routes from a set of stored routes as the current route is entered location by location. Route matching is made easy by comparing the current location to linear routes. An alert is generated when no matched routes exist. Preliminary experimental results show the proposed methods are effective and easy to use.

Mobile Data Protection Using Handheld Usage Context Matching

Handheld devices are easily lost because of their small sizes and high mobility. Personal data like addresses and telephone numbers stored in the devices are revealed when the devices are lost or used by unauthorized persons. This research proposes a novel approach for handheld data protection by using handheld usage context matching. The proposed system consists of five components: (i) usage data gathering, (ii) usage data preparation, (iii) usage pattern discovery, (iv) usage pattern analysis and visualization, and (v) usage pattern applications. Handheld usage context data is collected before applying this method. The usage context is then used to check device usage. When an unusual usage context such as an unlawful user trying to access the handheld data is detected, the device will automatically lock itself down until an action, like entering a password, is taken. Experimental results show this method is effective and convenient for mobile data protection.