Cap Ones Iscram09

ISCRAM 2009 12/05/2009Gothenburg

CAP-ONES: An EmergencyNotification System for all

Alessio Malizia, Pablo Acuña, Teresa Onorati, Paloma Díaz,Ignacio Aedo

DEI LabDepartamento de Informática

Universidad Carlos III de Madrid, Spain{amalizia, pacuna, tonorati, pdp}@inf.uc3m.es,

[email protected]


Contents• Introduction• The SEMA4A Ontology• CAP-ONES

– Alerts– Profiles– Ontology Queries– Notifications

• Use Cases• Conclusions


Introduction• Emergency Notification Systems (ENS) for

communicating alerts in emergency situations.• Accessible notifications depend on peopleʼs abilities,

device characteristics and kind of emergency.• ENSs must adapt notifications to the most appropiate

media and format to support alerts for all.


Introduction (2)• Keynote talk today - Martha Grabowski

– Situational awareness• Situational disabilities

– Smoke during a fire can cause low vision problem• Accessibility (be aware of the alert)


Introduction (3)• Common Alerting Protocol (CAP) provides a conceptual

framework to achieve interoperability.– FEMA Announces Intention To Adopt Common Alerting Protocol 1.1– Emergency Management Committee seeks feedback from CAP users– U.S. National Weather Service Starts Multi-Phase CAP Improvement Project– US Department of Homeland Security: "CAP Keeps Nation Steps Ahead of

Disaster"– U.S. Congressional Report Recognizes CAP

• We have developed a knowledge base in the form of anontology

• We have built a system that:– reasons over the ontology– to generate the most adequate notification– according to the user, the emergency and the device features.


The SEMA4A Ontology• Simple Emergency Alerts fo[u]r All interrelate knowledge

in the emergency alert space.• SEMA4A is formed by three main classes linked through

a number of relations among their subclasses:– EMEDIA: contains concepts and relations about emergencies

and media technologies.– WAfA: contains concepts and relations needed to model

organization, structure and navigation of information.– AccessOnto: contains information related to Web accessibility

guidelines, userʼs profiles and actions that users can perform.• Malizia, A., Astorga, F., Onorati, T., Diaz, P., Aedo I.: “Emergency Alerts for All: an Ontology

based Approach to Improve Accessibility in Emergency Alerting Systems”. ISCRAM 5thInternational Conference Information Systems for Crisis Response and Management, 2008, pp.197-207.


The SEMA4A Ontology(2)Internet communication capabilities

Guidelines for deafblind


CAP-ONES• Common Alerting Protocol-based Open Notification

Emergency System, takes as basis SEMA4A ontologyand creates automatically personalized emergencynotifications.

…PARSING… ...PROCESSING...

SMSMMS

.

.

.

CAP Alerts

Profiles

Information parsingSEMA4A Ontology

Personalized notifications


CAP-ONES: inputs


SMSMMS

.

.

.

CAP Alerts

Profiles




CAP-ONES: Alerts• The emergency information is extracted from a CAP alert,

containing data like source, status, description, location,etc.

• Our prototype supports entering information via a Webpage, importing a CAP file or providing an URL pointingto one.


CAP-ONES: Profiles• The profiles file is the second input, including:

– Personal and Contact Information: name, address, etc.– Abilities: level of userʼs abilities (low, medium or high) in 6

categories (Cognitive, Hearing, Coordination, Tactile Sensation,Visual and Colour) codified in the ontology.

– Devices: a selection from a list of possible devices according tothe userʼs abilities, extracted from relations defined in theontology.


CAP-ONES: queries


SMSMMS

.

.

.

CAP Alerts

Profiles




CAP-ONES: Ontology Queries• Using the information from the profiles and the

emergency, SPARQL queries are executed on theontology:1. Retrieve the media that can be used by this emergency, using

the RDF property mayUse defined in SEMA4A.2. Retrieve what can be communicated through the media

obtained in step 2, using the RDF property can-communicatedefined in the ontology.

3. For each profile, obtain the media and devices that the user isable to manage depending on his/her abilities, previouslyobtained when entering the profile with the system.


CAP-ONES: notifications output


SMSMMS

.

.

.

CAP Alerts

Profiles




CAP-ONES: Notifications• Using the set of media and the set of devices that can be

used for the emergency defined in the CAP message,• Using the media and devices for each profile,• A final result set per profile is computed with the

intersection of these sets.• The notification is created by selecting the appropiate

content from the CAP alert, according to the media anddevices contained in the final result set.


Use Cases• Profile: using our system, a “low hearing” and a “blindness”

profile are entered, corresponding to the Deafness andBlindness class in our ontology.

• Emergency alert: a CAP XML message of an earthquake(class Earthquake in SEMA4A), including an auxiliar imageresource, and a geographical location of the emergency.

User abilities Selecting Devices


Use Cases (2)• Our system response is the following:

– Profile Deafness may use the following media classes (frominterface): [Figure, Text, mms, email, sms, vibration]

– Profile Blindness may use the following media classes: [Sound, Text, mms, email, sms, vibration]

– By querying the ontology, class Earthquake may use the media:[tv, radio, mobile_phone, phone, internet];and can-communicate the following set: [Video, Sound,multiple_languages, Figure, Text, mms, email, sms, vibration]

– Making an intersection with the sets, we get that e-mail, sms andmms are feasible to send notifications in this situation.

– Personalization: the CAP message contains an image and ageographical location (that can be shown using figures), we validatethis against our results. Since the emergency can-communicateFigures, as well as the Deafness profile, so a link to Google Maps™pointing to the location can be added to an e-mail; however, this isnot added in the Blindness profile, since Figures can not be used.


Use Cases (3)

Deafness Profile

Blindness Profile


Use Cases (4)• Our system also allows sending SMS/MMS to users.

Since the emergency CAP message contained animage, we may use a MMS to send it as an auxiliarresource. As mentioned before, the Deafness profile canuse MMS and Figures; however, the Blindness profileonly allows sms.

MMS sent toDeafness Profile

SMS sent toBlindness Profile


Conclusions• We presented CAP-ONES, a prototype for adapting alert

notifications to different kinds of users depending on theirabilities, the kind of emergency and the devices they canaccess.

• Our system supports the CAP standard, for allowinginteroperability with existing systems, e.g. EDIS(http://edis.oes.ca.gov/)

• Our system is based on SEMA4A ontology, that has proven tocontain valid and useful concepts and relations.– Validate by experts (Sidar, Civil Protection)

• This is a first step towards a system that could automaticallyadapt alert notifications and interoperate with other systems.


Future Works• Extend the ontology to other classes

– Situational disabilities, new kind of devices• Test with scenario and users


SPARQL Query• SPARQL query example: to obtain the devices and media for a

specific emergency class (e.g. Earthquake):

• This query takes the earthquake class, a subclass of Emergency,and retrieves the object values from property mayUse, which getsthe media that can be used for this emergency; from these media,the query retrieves the object values from property can-communicate in order to obtain the tools that can be used to sendthe notification.

SELECT DISTINCT ?mayUse WHERE {:earthquake rdfs:subClassOf ?restrict ;rdfs:subClassOf :Emergency .?restrict rdf:type owl:Restriction ; owl:onProperty:mayUse ;owl:someValuesFrom ?emgMayUse .?emgMayUse rdfs:subClassOf ?emgMayUseClass .?emgMayUseClass owl:onProperty :can-communicate ;owl:someValuesFrom ?mayUse }


Ontology Validation• The expert on accessibility evaluated, totally, 155

elements extracted from our ontology.• Results were:

– • Coverage 91% (have all the lexons to be discovered actually beendiscovered?)

– • Precision 84% (are the lexons making sense for the domain?)– • Accuracy 79% (are the lexons not too general but reflecting the

important terms of the domain?).

• The expert on emergency evaluated, in total, 265elements extracted from our ontology.

• Results were:– • Coverage 66%– • Precision 65%– • Accuracy 45%

Cap Ones Iscram09

Business

emergency information

cap users

cap file

cap message

kind of emergency

emergency situations

common alerting protocol

personalized notifications