a conceptual framework for development of context-aware ... · a conceptual framework for...

. . . . . .

.

......

A Conceptual Framework for Development ofContext-aware Location-based Services on

Smart-M3 platform

Eldar Mamedov

P.G. Demidov Yaroslavl State University

Eldar Mamedov A Conceptual Framework for Development of Context-aware LBS

. . . . . .

.. Location-based Services

Location-based services (LBS) are services oriented at mobile usersthat take current position of the user into account whenperforming their tasks..LBS and context..

......

LBS take into account only a particular part of the context, thelocation of the user.

If LBS will consider other aspects of a user context into accountthey could provide more relevant information and appropriateservices for customers.

Examples of context-aware LBS:Routing for cars considering traffic conditionsRouting for people considering the public transport location


. . . . . .

.. Framework for Context-aware LBS

The framework is designed to provide relevant objects from thedatabase to the user taking his/her preferences and context intoaccount..Framework foundations..

......

Smart-M3 platformPreference SQL system

.Smart-M3 platform..

......

Semantic information broker (SIB)The core component that stores shared semantic data.Knowledge processors (KPs)The active parts of a system which can provide, modify andquery data.


. . . . . .

.. Preference model

.Preference definition1..

......Preference is considered as a strict partial order on the domainvalues of attributes of the database relation..Preference constructors..

......

Base constructors:BETWEEN, POS, LESS THAN, LOWEST, . . .NEARBY, WITHIN, ONROUTE, . . .

Complex constructors:Pareto preference (⊗)Prioritization preference (&)

1W. Kießling, “Preference queries with SV-semantics.” in COMAD, vol. 5, 2005, pp. 15–26.Eldar Mamedov A Conceptual Framework for Development of Context-aware LBS

. . . . . .

.. Context model within framework

.Context definition..

......Context is considered as a set of properties related to the user andrepresenting his/her state.

.Context types..

......

Static user contextContains user description that rarely changes. Usually staysthe same throughout series of requests.Dynamic user contextContains user description that frequently changes. Could bedifferent for each request and may change during the requestprocessing.


. . . . . .

.. Framework overview

.Different kinds of input..

......

Request from the client serviceCorresponding user’s contextUser profile

.Main roles of the framework..

......

Client KPContext-aware preference term generator KP(CAPTGenerator KP)Preference query executor KP (PQE KP)


. . . . . .

.. Data flow of the framework


. . . . . .

.. Data structures for request presentation and processing

UserContext

. . .

UserRequest

objectType : stringprocessed : boolean

User

PreferenceTerm

ResultItem

position : intDBObject

UserProfile

UserContext

. . .

UserProfileItem

objectType : string

relatesTo containsDynamicContext

consistsIn

resultsInconsistsIn

hasProfile

hasStaticUserContext

consistsIncontains


. . . . . .

.. Preference term representation

BetweenPreferenceTerm

lower : numberhigher : number

AttributePreferenceTerm

attribute : string

POSPreferenceTerm

preferredSubset : set

PreferenceTerm

ParetoComposition PrioritizedComposition

contains rightOperand,leftOperand


. . . . . .

.. Client KP

Adds an objects that describe the user: User, UserContextand UserProfileForms a request by adding a UserRequest, UserContext andPreferenceTerm objects to the SIBSubscribes for the pattern (UserRequest ID, ’pqe:processed’,’true’) to detect the moment when request is processedDisplays results to the user when the request is processed


. . . . . .

.. CAPTGenerator KP

Adds a ContextAwareGenerator objectSubscribes for the pattern (ANY, ’rdfs:class’,’pqe:UserRequest’) to track the addition of particularUserRequest objectsWhen the UserRequest object is added and matches the KPspecifications it reads dynamic context associated with therequest and static context bound to the user and generatespreference terms out of these dataWhen the KP prepares the preference term from the context itadds a ProcessedRequest object containing this resulting termand associate it with the UserRequest


. . . . . .

.. PQE KP

Subscribes for the pattern (ANY, ’rdfs:class’,ContextAwareGenerator) to track the addition and removal ofContextAwareGenerator objectsSubscribes for the patterns (ANY, ’rdfs:class’, UserRequest)and (ANY, ’pqe:generates’, ANY) to track the addition ofUserRequest and ProcessedRequest objectsCombines the user request, context and profile preferenceterms together by the following prioritization order<user request> & <user context> & <user profile>2

Executes the resulting preference term on the object database,adds resulting list of objects to the corresponding UserRequestobject and assigns ’true’ to the value of the processedproperty of the UserRequest object

2P. Roocks, M. Endres, H. Alfons, W. Kießling, and S. Mandl, “Design and implementation of a framework forcontext-aware preference queries,” Journal of Computing Science and Engineering, vol. 6, no. 4, pp. 243–256, 2012.


. . . . . .

.. Conclusion

Presented a conceptual framework for development ofcontext-aware location-based servicesDesigned an architecture of the framework based on theSmart-M3 platformDescribed main components and data structures of theframework


a conceptual framework for development of context-aware ... · a conceptual framework for...

Documents