適境化行為模式探勘於行動車輛資訊服務中之事件預測

Contextualized Behavior Mining for Event Prediction in Telematics

10.6382/JIM.200507.0039

苑守慈(Soe-Tsyr Yuan)；謝宜錚(Yi-Cheng Shia)

行動車輛資訊服務 ； 時空資料庫 ； 類神經網路 ； 屬性導向歸納法 ； Telematics ； spatial-temporal databases ； neural network ； attribute-oriented induction

資訊管理學報

12卷3期（2005 / 07 / 01）

39 - 66

繁體中文

Telematics乃是結合通訊及資訊，主要概念是指汽車的通訊及資訊技術，在生活應用上是指以汽車作為與外在環境資源的互動，結合無線通訊技術將資訊及多媒體等內容作單向或雙向傳輸的系統。而Telematics之內容與服務提供部分則扮演著服務是否獲得消費者青睞的關鍵角色。本研究嘗試結合Telematics與車險業者以提出一個能獲得消費者青睞的Telematics應用服務創新的範例。本研究藉由對車險業者其顧客交易資料，考量「人」、「地點」、「時間」此三維度，來分析、學習並預測顧客的行為模式，進而進行適境化、個人化之Telematics資訊提供。例如，當顧客於特定時間行經特定的區域時，Telematics對顧客進行資訊提供，而提供的資訊服務包含三種類別：失竊、刮傷、緊急救援相關警告及內容訊息。本研究亦提出一個新的適境化行為模式探勘模式以達到適人、適地、適時之Telematics資訊提供。

Telematics is a combination of automotive telecommunications and information processing that offers to drivers and their passengers useful wireless voice and data (e.g., location-specific security, information, and in-vehicle multimedia contents/services). This paper presents a new telematics application that is of collaboration with the auto insurance industry. Taking into account the three aspects (person, location, and time), the application analyzes, learns, and predicts a customer’s contextualized behavior, timely furnishing the customer with the important contextualized information in prevention of mishap. A combination of data mining and mobile location-based services is applied in this application so as to provide auto customers with certain precautious information that is personalized and contextualized. Such precautious information is regarding steal, scrape, and emergency when customers passing through certain locations at certain time. The main contribution of this paper rests on the presentation of a novel telematics application (using a new modeling of two learning methods) that embodies the distinguished value of mishap event prevention.

1. C. F. Herot(1980).Spatial Management of Data.ACM Trans. on Database Systems,5(3),493-514.
2. Chatschik Bisdikian,Isaac Boamah,Paul Castro,Archan Misra,Jim Rubas,Nicolas Villoutreix,Danny Yeh,Vladimir Rasin,Henry Huang,Craig Simonds(2002).Intelligent Pervasive Middleware for Context-Based and Localized Telematics Services.ACM Mobile Commerce Workshop.
3. D. J. Peuquet(1994).It`s About Time: A Conceptual Framework for the Representation of Temporal Dynamics in GIS.Annals of the Association of American Geographers,84,441-461.
4. D. J. Peuquet,Niu Duan(1995).An Eevent-Based Spatiotemporal Data Model (ESTDM) for Temporal Analysis of Geographical Data.International Journal of Geographical Information systems,19(1),7-24.
5. Jiawei Han,Yandong Cai,Nick Cercone(1992).Knowledge Discovery in Databases: An Attribute-Oridented Approach.Proceedings of the 18(superscript th) VLDB Conference Vancouver.
6. Jiawei Han,Yongjian Fu(1996).Exploration of the Power of Attribute-Oriented Induction in Data Mining.Advances in Knowledge Discovery and Data Mining.
7. May Yuan(1996).Temporal GIS and Spatial-Temporal Modeling.Third International Conference/Workshop on Integrating GIS and Environmental Modeling,New Mexico:
8. Soe-Tsyr Yuan,Eva Tsao(2003).A Recommendation Mechanism for Contextualized Mobile Advertising.Expert Systems with Applications,24(4),399-414.
9. Soe-Tsyr Yuan,Kai-Hsiang Peng(2004).Location Based and Customized Voice Information Service for Mobile Community.Information Systems Frontiers,6(4),297-311.
10. Java 2 Platform Enterprise Edtion Specification
11. Tamas Abraham(1999).Knowledge Discovery in Spatio-Temporal Databases.Ph. D. dissertation.
12. Wei Lu,Jiawei Han,Beng Chin Ooi(1998).Discovery of General Knowledge in Large Spatial Databases.Proceedings of Far East Workshop on Geographic Information Systems.