臺灣博碩士論文加值系統

近年來由於全球各地之氣候異常變化下，天然災禍肆虐導致生命及資源受極大威脅，示警資訊便成為人們對於災害發生來臨前後唯一可依靠之訊息，至今已廣泛地運用於各式各類危急情形之應用中，可透過電視、廣播及網際網路等不同接收方式掌握警戒之最新狀態，以降低生命及財產之損害。由於災害伴隨之示警種類多元，需依靠政府之分工，將合適之種類指定特定之權責單位發布。然而不同權責單位各司其職，採用各自訂定之標準作業規範記錄及發布示警資訊，造成來源不同且目的不一，獲取豐富之示警內容反而成為難以統合之隱憂，難以一次掌握當下發生之所有情形，導致異質性資訊之產生。因此，如何整合示警資訊，提供完整與即時之示警，並可讓後端之所有使用者正確解讀，甚至發展可結合應用系統與行動裝置之機制，便成為重要之挑戰。
本研究之主要策略為提出示警必須涵蓋核心資訊之開放式架構，採用5W1H分析法訂定核心資訊，作為示警必須提供最低需求之必要項目，並檢視不同標準及類型之示警資訊，並針對必要項目進行標準化之內容配套。基於5W1H之核心資訊下，須討論理想之整合運作，權責單位依據各種類之示警須遵循對應之發布模式，於正確之時刻給予合適之示警類型，規劃上下架機制之流程，提供所有示警最新之狀態，並於地理資訊平台展示具有空間特性之示警資訊。本研究另針對如網頁及手機等不同接收模式，探討必須供應之核心資訊內容，因應各模式之特性及限制，滿足使用者接收示警之需求，且由於近年志願者提升自願性示警之供應，進而分析此類示警之供應內容及因應模式，整合多項示警之來源，以達資訊之互操作性(Interoperability)。
本研究之主要貢獻為基於開放架構下提出示警必須涵蓋之核心資訊，透過標準化之項目可因應不同種類及模式之描述需求，提供之示警資訊將可於共同之基礎上解讀、分析與決策；並以其為基礎，經由本研究之整合機制及標準化作業程序，根據適當之發布模式掌握所有示警發布、更新及解除之警戒狀態；且納入探討志願者提供自願性示警內容之具體描述，對於整體狀況之評估及災情狀態之確認皆有所助益，以彌補政府權責單位無法立即掌控之災情，以達到截長補短之效果。未來可發展整合領域知識與情境感知之應用服務於防災決策中；並利用物聯網之特性，納入適地性服務以善用手持裝置之便利，因應智慧應用之挑戰。並以此成果為基礎，可納入更多來源發布示警之整合，因應不同尺度之示警，以提昇防救災決策研訂之品質。

Summary

Natural disaster is a major threat to people living in Taiwan. Especially in emergency situations, sending timely alert messages to people under threat is a necessary mission to the government. However, heterogeneous alert messages issued by different responsible agencies often make the alert receiptants’ reading and interpretation work more difficult. This problem becomes even worse when alert messages are expected to distribute over the internet and processed by computers. Efforts for standardizing the content of alert messages, such as the Common Alerting Protocol (CAP), have been developed. Regardless simple or complex, we argue alert messages issued by anyone (e.g., government, NGO, or even citizens) must convey necessary alert information for interoperable parsing, reading, analysis and integration. Using the well-known 5Ws and 1H principle as major viewpoints, we summarized and proposed a list of core elements that must be included in any alert message standards. The analysis of various alert standards from different organzations supports our arguments. Despite the different design, the corresponding core elements of alert messages from different agencies can be successfully mapped and assimilated into the same warning system. This consensus list of core elements can also serve as a reference for agencies to examine if its alert design is appropriate. According to the basic properties of alerts, we further proposed four major alert publishing modes and scenario based on the core elements. Responsible agencies can then develop their SOPs based on the selected publish mode. Standardized alert messages are extremely useful for updating area under threats and issuing warning to the public. The interoperable common basis provides an even more powerful mechanism to improve the integrated use of both the alert content and actions even in the heterogeneous alert standard environment. Via timely and standardized alert messages, more lives can be saved and damage can be effectively reduced.
INTRODUCTION

Natural disasters are major threats to people living in Taiwan. Many people lost their lives due to the lack of timely notification about the coming threats. In 2001, Typhoon Toraji caused 214 death and huge damages in Taiwan. Government agencies began to develop emergency SOP to convey alert messages to the public and help those people in needs. In the same year, 24,000 people were evacuated from the disaster area before Typhoon Nari stroke. Timely alert message is valuable information that can really save lives. However, alert message are issued by different agencies according to their responsible duties in the governments. Since the responsible agencies independently develop their own alert contents, the reading and interpretation of alert messages surely become an interoperability challenge. In addition to the format issue, the dissemination mechanism between the responsible agencies and the general public also need special attention. Instead of specifically designed for humans, future mechanisms will no doubt largely depend on computer-based technology to improve the quality and efficiency of alert service. The design of alert messages thus must take open and standardizationan into consideration. Despite every domain or agency may have their own thinkings about the content of alert messages, we argued a commonly accepted set of elements must reside in every alert messages and can serve as the bisis for developing alert dissemination strategies. These elements are named “core elements” in this research, which are also the baseline of any design of alert messages. We start by introducing the well-known 5Ws and 1H principle to examine the related factors and determine the necessary core elements. We then compare various types of alert messages following different standard to verify the necessity and usage of the core elements and extensively analyze the appropriate standardized contents. With the changing status of threats, responsible agencies must accordingly issue appropriate message to the public. The scenario of executing alert/update/cancel operations for alert is examined for responsible agencies’ reference. The presentation of alerts must further take the different properties and limitation of publishing media into consideration, e.g., TV/Radio broadcasting often require reporters’ explanation, while location-based service requires users to interpret the content by themselves. We also consider the possible use of core elements to improve the integrity of alert information from non-expert volunteers. We believe the standardization of core elements in alert messages, regardless where it comes from, can help to parse and interpret alert messages to meet the requirements of all alert publishing platforms. The results of this research will serve as the foundation for building such an open and interoperable framework for disseminating alert messages.

MATERIALS AND METHODS

Alert information provided by different responsible agencies may cause heterogeneous problem. In this research, we adopt the 5Ws and 1H principle for designing the core elements as it presents the essential information that must be included in every alert message. The core elements are listed below and discussed as follows.
(1)WHO：”Who“ factor indicates the people who are related to alerts in a particular way.
 Publisher：The responsible agencies publish the alert to users. The name of the publisher should be provided in alerts to indicate its identity and the level of trust.
 Receiver：when needed, responsible agencies must specifically point out the people must receive the alert, e.g., people living in a particular region.
(2)WHAT：”What“ factor explain what the threat source is.
 Headline：Human readable description of the alert, its content must be concise and meaningful. .
 Identification：To uniquely identify an individual alert, the identification information is often authorized by the responsible agencies.
 Reference：Reference records the identification of the previous alert to establish the historical relationship between alerts.
 Type：As time goes by, the conditions of the threat may continuously change. Responsible agencies must maintain the correct status of the alert, so the alert type at least includes alert, update and cancel.
(3)WHY：”Why“ factor explains why the alert is published.
 Level of severity：The level of severity determined by the responsible agencies present direct information for further action, e.g., staying at home or immediate evacuation. The names of the levels should be standardized.
(4)WHERE：”Where“ factor describes the area under threat. Its description should be standardized and easy to be handled by GIS.
 Geometric presentation：The responsible agencies use polygon, point or circle to specify the sptial extent of the area under threat. Every geometric presentation must be coupled with a coordinate reference system.
 Geocode：Geocode is the name or the ID of the area under threat. Sometimes the alert mechanism must be able to transform geocodes into correspoinding polygons for further analysis, e.g., LBS.
(5)WHEN：”When“ factor specifies the valid time of an alert.
 Effective time：The time when the alert content becomes valid. People must take appropriate actions after the effective time.
 Expired time：The time when the alert is expected to end. If the threat does not end at the expired time, the responsible agencies must publish a new or update alert.
(6)HOW：”How“ factor suggests the actions the people must take after receiving the alert.
 Instruction：An umabiguous and straightforward instruction must be given to let users know how to react to the coming threast.

Table 1 shows the comparison between the core elements and the designed elements of the ssurveyed alert standards. The ideal scenario is all of the core elements are matched and can be processed in similar ways. The compared alert standards include Common Alert Protocol (CAP), Japan Meteorological Agency(JMA) – XML, Distribution Element and Commercial Mobile Alert Service and CMAC message. The results show most of the core elements are included and can be successfully matched.
During a hazard, it is necessary for the responsible agencies to continuously issue alerts to the people under threasts. After analyzing the possible scenarios between hazards and alerts, we proposed four publish modes for the responsible agencies to select (Figure 1)：
Figure 1. Published Mode
(1)Alert：Mode 1 is mainly used for alerts that has predetermined end time. This type of alerts automatically expires without issuing a cancellation message. Typical examples are earthquake alerts.
(2)Alert – Cancel：Mode 2 is similar to Mode 1, only it requires a cancellation message to confirm the status of the alert is terminated. This applies to situations where the end time cannot be accurately estimated when alert is issued. Typical examples include accidents in railways and high-speed rail.
(3)Alert-Update-Update-…：Similar to Model 1, alerts belong to Mode 3 have predetermined end time, but the content may change according to the situations in reality, so update of alert messages are necessary. Only few types of alerts belong to this category, e.g., the government first announce called off schools, then upgrade to the called off of both work and schools.
(4) Alert-Update-Update-…-Cancel：Mode 4 is similar to mode 3, only it requires a cancellation to terminate the alert. Most of the alerts belong to this category, where several status updates are necessary during hazards. According to their SOP, the responsible agencies will issue cancel messages to confirm the threats are over. Typical examples include typhoon, heavy rain, floods, road closed and debris flow.

Based on the proposed strategies, the alert systems can easily handle alerts from different agencies, correctly monitor the changing status and avoid missing alerts. This research includes a comprehsive design of workflow for all publish modes. In addition to the management issue, the alert message must adapt to the limitation of displayed media, such that only some core elements are chosen. Table 2 summarizes the information each media can provide based on the 5Ws and 1H principle.
For website on PC, all of the core element can be displayed with appropriate interface design. Cellphones using cell-broadcast system receive alerts based on users’ location. It therefore doesn’t need “WHERE” factor. Due to the limitation of screen and number of characters allowed, only parts of the core elements are chosen. Volunteer alert is often collected from non-professional citizens. Without appropriate technology aids, the alert information from citizens is often restricted to what they observe. However, its power comes from the crowdsourcing and the information become valuable if the factor of WHAT, WHEN and WHERE can be correctly processesd.

RESULTS AND DISCUSSION

In this research, we use Oracle database to manage the element of published alerts and use ArcGIS Server to display alerts on the map to analyze the continuously updated condition of disaster. Two tables specifically designed for the core elements are established. One table records the extracted alert information, and the other records the affected areas because one alert may correspond to more than one area. These two tables are connected via unique identifier. Following the above four publish modes, the developed process parses the contents of the alert, retrieve matched core elements and record in the Oracle database. We use Python to publish alerts in ArcGIS Server and control the display of the results through ArcGIS Javascript API on website. Two important findings are further discussed below:

 We use alerts of the Common Alerting Protocol and the Japan Meteorological Agency as examples to test the interoperable interpretation and management of alert messages. The result indicates the core elements and the XML-based open framework are advantageous for handling alerts following different standards.
 Using typhoon as examples, the implemented prototype system demonstrates it can effectively and correctly monitor the various types of alerts in different phases. Users can use the fuction on the website to inspect the latest condition of each alert, for example, each layer contain different kind of alert. The system provides an interface for users to switch to alert of interests.

CONCLUSION

To integrate the various types of alerts from different sources, this research proposes the design of an open framework based on basic components of alerting messages. Following 5Ws and 1H principle, the proposed core elements include all of the essential imformation necessary for the fundamental operation of alert management. The core element serves as the basis to examine if an alert standard capable of supplying necessary information. Via the standardized core elements, different alert standards can be managed in an interoperabe way, as well as the alerts from providers can be correctly conveyed to users on a common background. In this way, the heterogeneous problem can be effectively resolved and alert standards are ingetrated accurately. The proposed four publish modes enable responsible agencies to adjust their current workflow or design new SOPs for publishing alerts that best fits the wanring requirements. The core elements also ensure the availability of essential alert information to adapt to users’ platform limitation. We also demonstrate the core elements can serve as the basis for improving the content of the volunteered alerts. The collaboration of government and citizens via standardized core elements can effectively integrate alert information in an accurate and timely fashion.

摘要.....I
TOWARDS THE ESTABLISHMENT AND OPERATION OF INTEROPERABLE ALERT INFORMATION FRAMEWORK.....III
致謝.....X
表目錄.....XIV
圖目錄.....XV
第一章、緒論.....1
1.1 研究背景.....1
1.2 研究目標.....3
1.3 研究流程.....5
1.4 論文架構.....8
第二章、文獻回顧.....9
2.1 示警流通之發展.....9
2.1.1 初期示警之運作模式.....10
2.1.2 示警資訊之流通.....12
2.1.3 網路示警之發展.....15
2.2 示警之產生與運作模式.....19
2.2.1 示警內容之討論.....19
2.2.2 示警分工多樣性.....22
2.2.3 群眾外包.....27
2.3 示警資訊標準化.....32
2.3.1 示警資訊之標準.....32
2.3.2 各國於示警標準下之運作.....39
2.3.3 示警系統之應用發展.....44
第三章、示警內容描述架構.....50
3.1 示警流通之情境分析.....50
3.2 示警資訊之分析.....63
3.2.1 5W1H分析法 .....63
3.2.2 示警核心資訊.....71
3.2.3 示警核心資訊之關聯性.....77
3.3 示警標準之內容分析.....79
第四章、示警資訊運作模式.....86
4.1 單一示警資訊之運作.....86
4.1.1 示警類型之使用原則.....87
4.1.2 示警發布模式.....90
4.1.3 示警之接收模式.....97
4.2 示警流通機制設計.....104
4.2.1 上架流程之機制.....105
4.2.2 下架流程之機制.....109
4.3 示警資訊之標準化課題.....110
4.4 自願性示警之供應.....116
4.4.1 自願性示警之內容.....116
4.4.2 自願性示警之運作.....120
第五章、系統實作與情境模擬 .....122
5.1 實作系統測試.....122
5.1.1 系統運作環境.....122
5.1.2 系統運作測試.....126
5.2 示警之互操作性.....135
5.2.1 共通示警協定.....136
5.2.2 日本氣象廳之XML規範.....138
5.3 示警情境之運作及應用.....142
5.3.1 單一示警之運作情形.....142
5.3.2 多示警之情境模擬.....147
第六章、結論與建議.....159
參考文獻.....162

大紀元. (2003). 歷史故事：蘇武牧羊. from http://www.epochtimes.com/b5/3/8/21/n363126.htm
中華文化總會. (2012). 中華語文知識庫 -示警. from http://chinese-linguipedia.org/clk/search/%E7%A4%BA%E8%AD%A6/98688/193307
方鵬程. (2007). 西方戰爭傳播演進之歷史分析: 以 [克力斯瑪],[傳播科技],[媒體在戰爭的角色] 的探討. 復興崗學報(89), 161-186.
氣象預報警報統一發布辦法 (2012).
交通部公路總局. (2013). 防災先預警，人車平安行-公路防災預警機制.
交通部公路總局. (2010). 交通部公路總局封橋封路標準作業程序(SOP).
土石流災害預報與警報作業手冊 (2012a).
各類災害警戒顏色燈號訂定原則 (2012b).
土石流防災疏散避難作業規定 (2010).
何德功, ＆ 劉浩遠. (2011). 日本媒體怎樣報導自然災害. 一五一十週刊(4).
吳杰穎, 鄭春發, ＆ 鄭國泰. (2006). 考量減災觀點之都市計畫通盤檢討防災規劃: 以淡海新市鎮特定區為例. 環境與世界(14), 71-92.
李信漢, ＆ 杜綺文. (2007). Web 2.0, 麥克魯漢知多少? 新聞學研究(92), 183-192.
林元輝. (2000). 日本阪神震災的傳播檢討. 新聞學研究(62), 4.
林李耀. (2007). 颱風災害與應變. 科學發展(410), 6.
林幸蓉. (2002). 電話型制的演變史. 台灣工藝(11), 8.
林照真. (2009). 電視媒體與災難管理-災難新聞的倫理困境. 廣播與電視(31), 55-79.
秦新龍、黃文亭. (2011). 百年氣象特展─臺灣氣象百年回顧之初探. 8.
陳佳君, 陳寧, 林巧璉, ＆ 蔡紀眉. (2010). 台灣報紙災難事件新聞報導研究——以四大報之八八水災新聞為例. 中華傳播學會年會, 35.
陳怡良. (2009). 民國98年颱風調查報告－第8號莫拉克(Morakot)颱風(0908): 中央氣象局氣象預報中心.
黃文亭. (2014). 颱風預報產品之演進. 科學研習月刊, 50(12), 7.
楊樺, ＆ 葉欣誠. (2012). 新聞媒體在天然災難事件中的角色和影響力. 危機管理學刊, 9(2), 63-70.
詹中原. (2004). 危機管理: 理論架構 (Vol. 5): 聯經出版.
廖振順. (2012). 地理課沒教的事_用Google Earth 大開眼界. 臺北市: 時報出版.
漢聲雜誌. (1985). 漢聲小百科 (Vol. 3). 臺北巿: 英文漢聲
劉崇堅, 張時中, ＆ 劉豐章. (2011). 考察日本總務省及NHK有關數位無線電視轉換、地震預警系統暨日本東京舉辦之有線電視及通訊論壇與展覽會報告書: 國家通訊傳播委員會.
気象庁. (2010). 気象庁防災情報XMLフォーマット1.1.
AGD. (2013a). Best Practice Guide for Warning Originators. Australia: Attorney General's Department.
AGD. (2013b). Emergency warnings. from http://www.em.gov.au/Emergency-Warnings/Pages/default.aspx
AGD. (2013c). Australia's Emergency Warning Arrangements (pp. 20). Australia: Attorney General's Department.
AGD. (2012). Emergency Data Exchange Language Common Alerting Protocol Australia Profile 1.0: Attorney-General's Department.
Bakillah, M., Li, R.-Y., ＆ Liang, S. H. (2015). Geo-located community detection in Twitter with enhanced fast-greedy optimization of modularity: the case study of typhoon Haiyan. International Journal of Geographical Information Science, 29(2), 258-279.
Basher, R. (2006). Global early warning systems for natural hazards: systematic and people-centred. Philosophical Transactions of the Royal Society of London A: Mathematical, Physical and Engineering Sciences, 364(1845), 2167-2182.
Binzel, C., ＆ Otting, M. (2003). Emergency broadcast message in a wireless communication device: Google Patents.
Botterell, A., ＆ Addams-Moring, R. (2007). Public warning in the networked age: open standards to the rescue? Communications of the ACM, 50(3), 59-60.
Buecheler, T., Sieg, J. H., Füchslin, R. M., ＆ Pfeifer, R. (2010). Crowdsourcing, Open Innovation and Collective Intelligence in the Scientific Method-A Research Agenda and Operational Framework. Paper presented at the ALIFE.
CanOps. (2012a). Canadian Profile of the Common Alerting Protocol Location References Beta 0.4: Canadian Public Safety Operations Organization.
CanOps. (2012b). Canadian Profile of the Common Alerting Protocol Beta 0.4A: Canadian Public Safety Operations Organization.
CanOps. (2012c). Canadian Profile of the Common Alerting Protocol Event References Beta 0.5: Canadian Public Safety Operations Organization.
Chow, C.-Y., Mokbel, M. F., ＆ Liu, X. (2006). A peer-to-peer spatial cloaking algorithm for anonymous location-based service. Paper presented at the Proceedings of the 14th annual ACM international symposium on Advances in geographic information systems.
Dawe, P. (2013). National Public Alerting System Common Look and Feel Guidance 1.0
de León, J. C. V., Bogardi, J., Dannenmann, S., ＆ Basher, R. (2006). Early warning systems in the context of disaster risk management. Entwicklung and Ländlicher Raum, 2, 23-25.
Drabek, T. E. (1991). The evolution of emergency management. Emergency management: Principles and practice for local government, 3-29.
Earle, P., Guy, M., Buckmaster, R., Ostrum, C., Horvath, S., ＆ Vaughan, A. (2010). OMG earthquake! Can Twitter improve earthquake response? Seismological Research Letters, 81(2), 246-251.
Estellés-Arolas, E., ＆ González-Ladrón-de-Guevara, F. (2012). Towards an integrated crowdsourcing definition. Journal of Information science, 38(2), 189-200.
FEMA. (2014a). Wireless Emergency Alerts (WEA) Fact Sheet: IPAWS Project Management Office.
FEMA. (2014b). IPAWS Toolkit for Alerting Authorities. In A. Johnson (Ed.), (pp. 52): IPAWS Program Management Office.
Gao, H., Barbier, G., ＆ Goolsby, R. (2011). Harnessing the crowdsourcing power of social media for disaster relief. IEEE Intelligent Systems(3), 10-14.
Geleta, B. (2012). Community early warning systems: guiding principles. Switzerland: The International Federation of Red Cross and Red Crescent Societies (IFRC).
Giles, J. (2005). Internet encyclopaedias go head to head. Nature, 438(7070), 900-901.
Golnaraghi, M. (2009). 2009 GLOBAL ASSESSMENT REPORT ON DISASTER REDUCTION:Thematic Progress Review Sub-Component on Early Warning Systems: World Meteorological Organization.
Goodchild, M. F., ＆ Glennon, J. A. (2010). Crowdsourcing geographic information for disaster response: a research frontier. International Journal of Digital Earth, 3(3), 231-241.
Griffin, T., ＆ Miller, M. K. (2008). Child abduction, AMBER alert, and crime control theater. Criminal Justice Review, 33(2), 159-176.
Gruteser, M., ＆ Grunwald, D. (2003). Anonymous usage of location-based services through spatial and temporal cloaking. Paper presented at the Proceedings of the 1st international conference on Mobile systems, applications and services.
Howe, J. (2006). The rise of crowdsourcing. Wired magazine, 14(6), 1-4.
Kraak, M.-J. (2004). The role of the map in a Web-GIS environment. Journal of Geographical Systems, 6(2), 83-93.
Morrow, N., Mock, N., Papendieck, A., ＆ Kocmich, N. (2011). Independent evaluation of the Ushahidi Haiti project. Development Information Systems International, 8, 2011.
Nishio, M., ＆ Mori, M. (2012). Web-based Delivery System for Disaster Prevention Information using a New JMA DPI XML Format and AMeDAS Data. ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 1, 4.
O'reilly, T. (2007). What is Web 2.0: Design patterns and business models for the next generation of software. Communications ＆ strategies(1), 17.
OASIS. (2013a). Emergency Data Exchange Language Situation Reporting 1.0.
OASIS. (2013b). Emergency Data Exchange Language Distribution Element 2.0.
OASIS. (2010). Emergency Data Exchange Language Common Alerting Protocol Version 1.2.
OASIS. (2009). Common Alerting Protocol, v. 1.2 USA Integrated Public Alert and Warning System Profile Version 1.0.
Okolloh, O. (2009). Ushahidi, or'testimony': Web 2.0 tools for crowdsourcing crisis information. Participatory learning and action, 59(1), 65-70.
Peary, B. D., Shaw, R., ＆ Takeuchi, Y. (2012). Utilization of social media in the east Japan earthquake and tsunami and its effectiveness. Journal of Natural Disaster Science, 34(1), 3-18.
Perez‐Lugo, M. (2004). Media uses in disaster situations: A new focus on the impact phase. Sociological Inquiry, 74(2), 210-225.
Petak, W. J. (1985). Emergency management: A challenge for public administration. Public Administration Review, 3-7.
PSC. (2014). National Public Alerting System. from http://www.publicsafety.gc.ca/cnt/mrgnc-mngmnt/mrgnc-prprdnss/ntnl-pblc-lrtng-sstm-eng.aspx
Sheth, A. P. (1999). Changing focus on interoperability in information systems: from system, syntax, structure to semantics Interoperating geographic information systems (pp. 5-29): Springer.
Sillem, S., ＆ Wiersma, E. (2006). Comparing cell broadcast and text messaging for citizen warning. Paper presented at the Proceedings of the 3rd International ISCRAM Conference.
Spencer, J. W., Seydlitz, R., Laska, S., ＆ Triche, E. (1992). The different influences of newspaper and television news reports of a natural hazard on response behavior. Communication Research, 19(3), 299-325.
Surowiecki, J. (2004). The wisdom of crowds: Why the many are smarter than the few and how collective wisdom shapes business, economies, societies and nations little. Brown ISBN 0-316-86173-1.
Whittaker, S., Jones, Q., ＆ Terveen, L. (2002). Contact management: identifying contacts to support long-term communication. Paper presented at the Proceedings of the 2002 ACM conference on Computer supported cooperative work.
Zook, M., Graham, M., Shelton, T., ＆ Gorman, S. (2010). Volunteered geographic information and crowdsourcing disaster relief: a case study of the Haitian earthquake. Available at SSRN 2216649.