=Paper=
{{Paper
|id=Vol-2215/paper16
|storemode=property
|title=Semantic-based Social Intelligence through Multi-Agent Systems
|pdfUrl=https://ceur-ws.org/Vol-2215/paper_16.pdf
|volume=Vol-2215
|authors=Michele Ruta,Floriano Scioscia,Giuseppe Loseto,Filippo Gramegna,Agnese Pinto,Eugenio Di Sciascio
|dblpUrl=https://dblp.org/rec/conf/woa/RutaSLGPS18
}}
==Semantic-based Social Intelligence through Multi-Agent Systems==
https://ceur-ws.org/Vol-2215/paper_16.pdf
Semantic-based Social Intelligence through
Multi-Agent Systems
Michele Ruta, Floriano Scioscia, Giuseppe Loseto, Filippo Gramegna, Agnese Pinto, Eugenio Di Sciascio
Polytechnic University of Bari - via E. Orabona 4, Bari (I-70125), Italy
{michele.ruta, floriano.scioscia, giuseppe.loseto, filippo.gramegna, agnese.pinto, eugenio.disciascio}@poliba.it
Abstract—Current technologies and market solutions are far and technical use cases are quite far from such levels of
from fulfilling the Ambient Intelligence (AmI) vision of simplified intelligence, automation and adaptivity. A limited flexibility is
people-environment interactions. Even though Despite recent possible, as devices are logically associated at the application
solutions based on Internet of Things (IoT) technologies provide
the needed infrastructure, most approaches suffer from inade- level by means of static profiles, defined during systems
quate levels of intelligence and autonomy. This paper proposes deployment. This is also the case of domotics –i.e., Home and
a novel semantic-based Multi-Agent System (MAS) framework Building Automation (HBA)–, one of the widespread examples
complying with the emerging Social Internet of Things paradigm of AmI for environmental control. In most established HBA
devoted to improve both automation and adaptivity: device agents standards, solutions are centralized and proprietary: changing
self-organize in social relationships, interacting autonomously
and sharing information, cooperating and orchestrating ambient possible configurations or introducing new devices typically
resources. A service-oriented architecture allows collaborative require the intervention of qualified practitioners. Even recent
dissemination, discovery and composition of service/resource “smart home” platforms introduced by IT companies still
descriptions. Decision and choreography capabilities of software depend heavily on manual configuration and provide only
agents leverage Semantic Web languages at the knowledge rudimentary levels of automation [3].
representation layer and a mobile-oriented implementation of
non-standard inferences for semantic matchmaking. Benefits of
the proposal are highlighted through an AmI case study in the This paper presents a novel MAS paradigm at the con-
field of Home and Building Automation (HBA). A comparison vergence of the Semantic Web of Things (SWoT) [4] and
with the state of the art is also provided. Social Internet of Things (SIoT) [5] visions. IoT devices act
Index Terms—Semantic Web of Things, Social Agents, Ambient as socially intelligent agents, capable of autonomous config-
Intelligence, Service Discovery
uration, coordination and orchestration. Interaction patterns
inspired by SNSs allow agents to establish relationships,
I. I NTRODUCTION
share information, exchange requests and services, in a dy-
The advent of Social Networking Services (SNSs) has had a namic, decentralized and collaborative fashion. Agents ex-
deep impact on how people communicate and interact. Starting ploit Knowledge Representation (KR) technologies borrowed
from personal user profiles containing general information, from the Semantic Web to express and circulate knowledge
typical elements of SNSs include: the capability to engage about themselves and the context they are dipped in. In
asymmetrical (e.g., follower/followee) or symmetrical (e.g., addition, the semantic-based matchmaking implemented in a
friendship, group) relationships among users; a personal log resource-efficient mobile engine [6], on a moderately expres-
(wall) to post text and/or multimedia items; the possibility to sive fragment of the Web Ontology Language (OWL2) [7],
mark (tag) contacts to draw their attention to a given item, supports the social intelligence through discovery, aggregation
as well as to append comments and reactions (e.g., like) to and ranking of available social entities. As agents acquire
elements published by other users. These basic primitives can new knowledge about their context, both their configurations
be combined to support several interaction models, granting and the environmental services evolve: the MAS becomes a
users high flexibility in the way they share information, social network, where individual device interactions produce
communicate, collaborate and search for resources of interest. emergent behaviors toward high-level goals, without requiring
Endowing autonomous agents with social capabilities can explicit user commands. The paper reports on a case study
transfer benefits of SNSs to Multi-Agent Systems (MASs), in the field of HBA: current approaches are compared with
particularly to complex, dynamic and loosely coupled ones. the one proposed here in order to assess possible benefits and
This is the case of Internet of Things (IoT) contexts for Am- evidence the added value of the proposal.
bient Intelligence (AmI) [1], where networks of lightweight
agents on highly heterogeneous mobile and embedded devices The remainder of the paper is organized as reported in what
provide context-aware, adaptable, unobtrusive and intelligent follows: after related work discussion in Section II, Section
support to users’ activities [2]. In AmI, the environment should III describes the proposed approach. An AmI case study is
adapt to changes in external conditions as well as users’ presented in Section IV to clarify the proposal, including a
personal preferences and requirements, even anticipating needs comparison with state-of-the-art technologies for IoT-oriented
and behaviors. Current solutions available for commercial HBA. Conclusion in Section V closes the paper.
96
II. R ELATED WORK TABLE I
N ETWORK ENTITIES AND SOCIAL FEATURES
The study in [8] observed that the majority of SNS users
Technical feature Social environment
access them regularly, as they find both enjoyment and use- Object / Device / Application Social agent
fulness. Higher numbers of connected users –and in particular Functional profile Service
complementary ones [8]– increase opportunities for finding Object pairing Social relationship
needed information and services. These benefits can also apply (friend/follower)
Object communication Social interaction
to social networks of objects, which work as independent Object configuration Distributed service discovery
agents and interact for information and resource/service shar- update/adaptation
ing. Object log Wall
Object command Post
One of the earliest examples of social object capabilities can Object reply Comment
be found in [9], a proposal aimed at distributed OWL Knowl- Functionality Tag & Like
edge Base management and reasoning. Upon connection to the activation/deactivation
network, embedded devices proactively exchanged information
in a handshake. “Requester” devices, endowed with reasoning
only manual configuration, but also changes to the reference
capabilities, allowed users to execute queries, which were
ontology. Also the proposals in [20], [21] relied on rule-
automatically distributed among requester’s “known” devices.
based reasoning, where the system state should fully match
Unfortunately, reasoning capabilities were curbed by the re-
rule conditions in order to trigger a rule. Unless only ele-
strictions of the adopted query language, limiting the practical
mentary rules are adopted, however, full matches are quite
interest of supported use cases.
rare in realistic scenarios, where entities are described by
The approach proposed in this paper is conceptually close
heterogeneous and often contradictory annotations. Finally, in
to [5], where SIoT has been envisioned as a social evolution
[22] a semantic Service-Oriented Architecture (SOA) enabled
of the Internet of Things, with agentified objects capable of
discovery and composition of semantic services. For greater
setting mutual relationships and exploiting them to exchange
autonomy and flexibility, in this paper the SOA paradigm
information and services, without requiring interactions with
has been coupled with a MAS of socially intelligent agents.
users or human-oriented SNSs. Conversely, earlier efforts such
The proposal extends the early conceptual and architectural
as [10] aimed to make objects aware of people’s social context.
elements introduced in [23].
Networks of socially intelligent objects were analyzed in [11],
by defining key metrics about nodes and links, adapted from III. A NETWORK OF SEMANTIC AGENTS
the literature on SNSs analysis. An ontology formalized the The proposed approach aims at agent coordination in pur-
definitions, and social objects could use them to manage their posely infrastructured environments and particularly in do-
policies, friends and reputation. A further step toward social motics scenarios through interaction paradigms inspired by
agency is object blogging [12], i.e., an object’s ability to social networks. Devices are fully enabled in sharing re-
self-describe autonomously on the Web or in a local area sources/services, making decisions, disseminating requests and
network to support intelligent interactions. This was previously gaining responses through a distributed peer-to-peer protocol.
explored in RFID contexts [13] and constitutes an evolution of Shared knowledge fragments about devices themselves, func-
proposals requiring human intervention [14], [15]. The work tional profiles and context are advertised via a decentralized
in [16] identified smartphones as means to put people back in service-oriented architecture. The social relationship and the
the loop of ubiquitous autonomous social MASs. Smartphones discovery models outlined hereafter integrate in a unified
are ideal tools for learning about their owners and context, in social agent space both single-purpose physical objects and
order to work as their digital counterparts, exposing dynamic applications deployed on multi-purpose devices.
personalized profiles in the social agent choreography. Sev-
eral works have already explored smartphones and wearable A. Framework and architecture for social agents
devices to model users’ activities, preferences and contexts Table I highlights basic correspondences of entities and
[17]. features in a generic AmI domain to the proposed social MAS
Semantic-based approaches are not uncommon in AmI environment. This applies particularly to domotics and HBA.
and particularly in domotics. Building automation ontologies Every object acts as a social agent: it exposes an individual
were used for system design and commissioning, device profile describing its general features (e.g., device type, lo-
description, data modeling and access, ambient control [18]. cation, hardware details) as well as the resources/services it
The ontology-based system in [19] delivered context-aware can provide through possible configurations. An agent is able
customized information to different kinds of users. Queries to become friend and/or follower of other agents. According
matched device and user descriptions in OWL while rules to the different kinds of interactions described hereafter, it
implemented temporal and extra-logical constraints, achiev- can write posts on either its wall or friends’ walls when its
ing overall capabilities similar to Complex Event Processing settings or capabilities change, and also when it produces new
(CEP) architectures. Nevertheless, integration appears as a or updated information after a context analysis. Each post
serious limitation, because installing new devices required not contains perceptions and events observed by the social agent.
97
In the proposed SOA-based MAS, it is considered as a request The proposed MAS complies with a range of different
for system reconfiguration through distributed semantic service scenarios and contexts, because it is inherently platform-
discovery, which can be exploited by: independent and general-purpose. All social features reported
in Table I can be modeled regardless of the particular
• sensor agents, such as a weather station, which can
application-layer communication protocol. Anyway, a first
observe the environment and share data but don’t have
implementation has been proposed in [24] based on the
actuation capabilities;
Constrained Application Protocol (CoAP) [25], a lightweight
• actuator agents, such as a lamp or a fan, which can react
Web of Things protocol for machine-to-machine interaction.
to environmental changes but have limited or no sensing
facilities: by reading posts, they become aware of current
conditions and activate/deactivate some services; B. Semantic-based social network interaction
• smart agents, endowed with both sensors and actuators: if
In the proposed framework, social agents are distinguished
a smart agent does not have all the capabilities needed to in two possible families: full ones, able to execute the inference
comply with the perceived changes, a discovery process tasks described above, and basic ones, endowed with low
is started to find peers providing further suitable services, memory and low (or no) computing capabilities, which can
as described in Section IV-A. only provide sensing/acting services, but cannot perform au-
Semantic annotations referred to ontologies in OWL2 [7] tonomous reasoning. A pair of agents can engage in two kinds
are used to express agent profiles, service descriptions and of social relationships. Through the bidirectional friendship
requests. Being formally grounded on Description Logics link, they can exchange both information and services. In
(DLs) semantics, they are both machine understandable and particular, they became able to: (i) read and write on each
human readable. In particular, this paper refers to the OWL2 other’s wall; (ii) request the friend’s service descriptions; (iii)
fragment corresponding to the ALN (Attributive Language activate or deactivate the friend’s services. When becoming
with unqualified Number restrictions) DL, which supports friend with a full agent, a basic agent can select it as semantic
standard and non-standard inferences with polynomial com- facilitator, i.e., reasoning helper. Conversely, an agent can
plexity [6]. follow another one if interested only in receiving updates
Decision capabilities of social agents are enacted through published on its wall, i.e., becoming an observer through a
a collaborative service/resource discovery. This process lever- unidirectional relationship.
ages semantic matchmaking, i.e., the task aimed at retriev- Following/friendship criteria are automatically verified by
ing and ranking the most relevant resources for a given means of a matchmaking process involving the device profiles.
request, where both requests and resources are satisfiable Two agents are good candidates for friendship if one or more
concept expressions w.r.t. a common ontology T . Classic of the following conditions are met: (i) strong co-location,
subsumption/satisfiability approach is extended here by means i.e., devices are placed in the same room/area; (ii) parental
of the Concept Abduction, Concept Contraction and Concept or co-ownership, i.e., they are from the same manufacturer
Covering [6] non-standard inference tasks in ALN : or belong to the same owner; (iii) co-working, i.e., they are
- Concept Contraction: if annotations of a request R and a able to cooperate closely as they share annotations referred
given resource S are not compatible (i.e., an explicit clash to the same ontology and provide functionalities related to the
arises from their logical conjunction), Contraction determines same activity (e.g., room lighting) or observed parameter (e.g.,
what part G (for Give up) of the request is conflicting with indoor temperature). On the other hand, a follower request is
S. If one retracts G from R, a concept K (for Keep) remains, more appropriate in case of: (i) weak or sporadic co-location,
which is a contracted version of R compatible with S. G such that information produced by an agent can still be useful
explains “why R and S are not compatible”; to other ones to characterize their own context, but at the same
- Concept Abduction: if R and S are compatible, but S does time they need/prefer to start independent discovery requests;
not satisfy R completely, Abduction determines what should (ii) no co-ownership; (iii) weak co-working relationship, i.e.,
be hypothesized in S in order to obtain a full match. The direct interactions would have low usefulness, because e.g., the
solution H (for Hypothesis) to Abduction explains “what is two agent profiles are incompatible w.r.t. a common reference
requested in R and not specified in S”. By computing penalty ontology, i.e., they are significantly different (note that even
metrics linked to G and H [6], Contraction and Abduction a follower relationship is inappropriate in case of profiles
further enable a logic-based relevance scoring of a set of referring to separate ontologies, as that implies agents belong
resources w.r.t. a certain request; to totally different domains, e.g., HBA and healthcare).
- Concept Covering: in AmI scenarios such as domotics, it is For a broader range of interaction patterns, the framework
often useful to compose multiple services/resources in order also permits being both a friend and a follower of the same
to satisfy a complex request. Given a request R and a set agent: this is useful in highly heterogeneous scenarios. In
of resource instances S = {S1 , S2 , ... , Sn }, Covering finds any case, a friendship/follow request can be rejected if the
out a pair hSc , Hi, where Sc ⊆ S contains resources whose above conditions are not verified or the maximum number
aggregation satisfies R as much as possible, while H is the of friends/followers has been reached w.r.t. processing and
(possible) remaining part of R not covered by concepts in Sc . memory limits of the invited agent. In practice, however, by
98
enlarging its social network an agent increases opportunities
A1 A2 A3
for useful cooperation, hence rejections should be infrequent.
Like in human-oriented SNSs, agents’ walls are the main writeInWall(P1) 1) write a post
knowledge sharing medium. The proposed framework supports covering(R1)
updatePost(P1) 2) local discovery
both push and pull models, exploiting the above relationships:
postInWall(P2) writeInWall(P2) 3) collaborative
• push: if agent Ai wants to receive updates from peer ACK
Aj automatically, it will ask to become a follower. If covering(U1) discovery
updatePost(P2)
accepted, follower Ai will be able to start a distributed postInWall(P3)
discovery session when it receives a notification of a new processing like A2
ACK
post or comment on the wall of the followed agent Aj ; notify() 4) receive
• pull: if Ai wants to access Aj ’s wall on demand, it readComment(P2) notifications
will ask to become a friend. By doing so, Ai will also comment
updatePost(P1)
automatically grant Aj access to its own wall. Then Ai
will perform semantic matchmaking if Aj writes a post
on Ai ’s wall, during a collaborative covering as reported Fig. 1. Sequence diagram for a distributed reconfiguration
in Figure 1.
Each agent will choose a model –or even both– based on its Friend
goals and strategies. These elements are relevant and conform
to the general behavior policy of the MAS, anyway they are
outside the scope of the paper. AC
When an agent detects an event (e.g., a change in internal or
environmental parameters) and conditions require adaptation – AS
i.e., modification to the functional configuration of itself and/or
of nearby devices– it will write a post on its wall. As these
trigger mechanisms are fundamentally domain-dependent and
application-oriented, the framework does not prescribe specific
solutions. In any case, the written post P will consist of a
pair hR, Li, where R is the request issued by the node – L
expressed as a semantic annotation w.r.t. a reference ontology– SC
and L is the like value. The like reaction to a post has been
mutuated from human-oriented SNSs, but in the proposed
approach it is a real value in [0, 1] instead of a Boolean Fig. 2. Case study scenario
value. It represents the coverage ratio of request R, as resulting
from Concept Covering in the collaborative service discovery
process. Specifically, if U is the uncovered part returned by steps 2) - 3).
the Concept Covering of R with a set of available services, 4) When Ai receives the notification of Pj , it reads the
the associated like value is computed as L = 1 − norm(U ) comment from the friend’s wall and appends it to Pi in order
norm(R)
using the norm on concept expressions described in [6]. An to update the status of the request. Finally, Ai updates the like
example of the whole process is in Figure 1, composed of the value accordingly.
The choice of friend(s) to call in the above step 3 basically
following steps:
depends on heuristic preference criteria, such as the number
1) When an agent Ai detects a reconfiguration is needed, it
and type of services exposed by the friend (known at friendship
writes a post Pi on its own wall. Li is initialized to 0.
establishment time), network latency or friend’s computational
2) If Ai is a basic device, go to step 3. Otherwise, Ai executes
resources.
the Concept Covering task on the local set of service anno-
tations S (Section III-A). Ai activates the selected services IV. C ASE STUDY: FROM OBJECTS TO AGENTS FOR
and adds a comment Ci to Pi as a pair hUi , Ti i, where Ui A MBIENT I NTELLIGENCE
is the uncovered part of Ri and Ti tags the selected local The case study presented here would clarify the social
services/resources. Moreover, the value of Li is updated as and collaborative potentialities of the proposed MAS frame-
per the above formula. work. To this aim, a specific scenario is targeted: the self-
3) If Ri is not completely covered, Ai selects a friend Aj and orchestration capability of agentified home devices allows to
writes a post Pj =hRj , Lj i) on its wall. Particularly, if Ai has evidence the AmI capabilities of the above approach.
executed step 2, Rj is set to the uncovered part Ui , otherwise
Rj is equal to Ri and Lj is 0. Writing Pj on the friend’s A. Illustrative example
wall automatically implies that Ai must be notified when a Figure 2 depicts the reference testbed recalling the case
comment is added to the post. Aj recursively executes the study; a house contains a social network of semantic-enabled
99
AS_Request: (detectsOutdoorLuminosityCondition some) Lamp_On: (detectsOutdoorLuminosityCondition some)
and (detectsOutdoorLuminosityCondition only and (detectsOutdoorLuminosityCondition only
LowLuminosityCondition) and (detectsIntrusionEvent LowLuminosityCondition) and (detectsIntrusionEvent
some) and (detectsIntrusionEvent only some) and (detectsIntrusionEvent only
IntrusionEvent) and (detectsOccupancyCondition some) IntrusionEventForLamp)
and (detectsOccupancyCondition only (not
OccupantPresence)) Lamp_M edium: (detectsOutdoorLuminosityCondition
some) and (detectsOutdoorLuminosityCondition only
Fig. 3. Request posted by the alarm system AS MediumLuminosityCondition) and
(detectsOccupancyCondition some) and
(detectsOccupancyCondition only OccupantPresence)
F ull_Close: (detectsPrecipitationCondition some) and
(detectsPrecipitationCondition only Rain) and Lamp_Of f : (detectsOutdoorLuminosityCondition some)
(detectsWindCondition some) and and (detectsOutdoorLuminosityCondition only
(detectsWindCondition only StrongWind) and HighLuminosityCondition) and
(detectsIntrusionEvent some) and (detectsOccupancyCondition some) and
(detectsIntrusionEvent only (detectsOccupancyCondition only (not
IntrusionEventForShutter) and OccupantPresence))
(detectsOccupancyCondition some) and
(detectsOccupancyCondition (not OccupantPresence)) Fig. 5. Dimmer lamp L service annotations
Half _Close: (detectsPrecipitationCondition some) and
(detectsPrecipitationCondition only (not Rain)) and
(detectsWindCondition some) and listed in Figure 4, Concept Covering selects only
(detectsWindCondition only ModerateWind) the Full Close service, provided by SC: this is
Open: (detectsPrecipitationCondition some) and basically due to commonality with the request
(detectsPrecipitationCondition only (not Rain)) and of concepts (detectsOccupancyCondition
(detectsWindCondition some) and some) and (detectsOccupancyCondition
(detectsWindCondition only LightBreeze) and
(detectsOutdoorLuminosityCondition some) and (notOccupantPresence)) (service descriptions
(detectsOutdoorLuminosityCondition only provided by the air conditioner are not reported because it
HighLuminosityCondition) does not offer any useful feature). AS comments its post
Fig. 4. Shutter controller SC service annotations including both a tag to the Full Close shutter service and
the uncovered part of the request. In order to further cover
the post, AS can select one of its friends and forward
agents embedded in the following devices: an alarm system the uncovered part. Since SC has provided the highest
(AS), a rolling shutter controller (SC), an air conditioner (AC) contribution in the first covering step, AS posts on SC’s
and a dimmer lamp (L). The blue arrows in Figure 2 specify wall the OWL2 annotation of the uncovered part, reported in
the existing friendship relations between the above agents. Figure 6. When SC receives the message, it recursively starts
According to the criteria suggested in Section III-B, the a covering process, which involves the services exposed by
agents set friendship relations because they are in the same its friend L (Figure 5). The Covering inference task selects
location and share functionalities useful to improve comfort the Lamp On service, which completely covers the remaining
or security in the house. Not all agent pairs are friends: in part of the initial request. SC therefore comments the post on
particular, Figure 2 shows L befriends SC only. Besides, each its wall by tagging the activated service and updating the like
agent has embedded sensing and/or actuating capabilities and value to 1. Further agents do not need to be involved, as the
exposes a set of functional profiles to its friends. request is fully satisfied. Finally, AS receives a notification of
Let us suppose it is evening and AS detects an intrusion the comment to its post on SC’s wall, it reads the comment
while there is nobody in the house. AS writes a new post and sees the initial request has been completely fulfilled. As
on its wall, representing what it has sensed as an OWL2 a consequence, it updates the like value of the post on its
annotation. Figure 3 shows a formalization of the post in wall and the discovery process stops. The house has changed
OWL2 Manchester syntax [26]. Service requests and descrip- its configuration by closing shutters and switching the lamp
tions are expressed w.r.t. the reference ontology (not reported on reacting to the intrusion alert.
due to space constraints), by specifying the context conditions
suitable for the activation of a given service. Then AS starts AS_Req_U ncovered: (detectsOutdoorLuminosityCondition
some) and (detectsOutdoorLuminosityCondition only
a Concept Covering process using the content of the post as LowLuminosityCondition) and (detectsIntrusionEvent
request, while services are taken from AS’s cache of available some) and (detectsIntrusionEvent only
functionalities exposed by all its direct friends, i.e., SC and IntrusionEventForLamp)
AC. Freshness of cache entries is checked via preliminary Fig. 6. OWL2 annotation of the uncovered part of AS Request
conditional requests: a service annotation will be retrieved
again only if it has been updated, otherwise AS can directly It is useful to point out that the Intrusion class was
use the cached copy. This procedure guarantees the covering defined as more specific than both IntrusionForLamp and
task is performed using the latest descriptions of all available IntrusionForShutter, i.e., it should require services
services. both from a lamp and a shutter controller. Such a model-
According to the semantic service descriptions ing pattern allows activating functionalities (Full Close and
100
TABLE II
C OMPARISON WITH CURRENT I OT- ORIENTED FRAMEWORKS FOR HBA
Features KNX IoT IzoT Platform Dog Gateway Eclipse SmartHome Proposed Approach
Home Area Network multi-protocol multi-protocol multi-protocol
EIB/KNX LonTalk
reference protocol over HTTP over HTTP over CoAP
Network architecture centralized centralized centralized centralized full P2P
via LonBuilder Domain-Specific
Network/devices XML configuration autonomous social
via ETS software software or XML Language (DSL)
configuration files agents configuration
configuration files configuration files
edit network/devices edit network/devices edit network/devices edit network/devices agents
Add/remove devices
configuration configuration configuration configuration self-configuration
Multi-protocol node acting smart agents acting
KNX gateway IzoT gateway Dog gateway
communication as gateway as gateways
static, defined during static, defined during dynamic, based on static, defined during dynamic, based on
Device binding
network configuration network configuration device profile device configuration friendship relationships
Scenarios static, defined during static, defined during dynamic, exploiting static, based on an ECA dynamic, exploiting
configuration network configuration network configuration rule-based reasoning rule engine non-standard inferences
yes, through
Service composition no no no no
distributed covering
proprietary proprietary, XML-based
Message data format OWL 2 XML OWL 2
(KNX specs.) (LonTalk specs.)
Standardized KNX IoT WebSocket and CoAP
HTTP RESTful API HTTP RESTful API
framework interface Web Services HTTP RESTful API RESTful interface
Lamp On) of different devices that are fired when the same here fully complies with resource-constrained scenarios (by
event is detected. supporting a P2P architecture and lightweight protocols such
The above example has been kept simple for the sake of as CoAP). Another distinguishing feature is a certain expres-
clarity, with relatively short service annotations and purely re- siveness in the possibility of device description and modeling
active MAS behavior. Notwithstanding, the adopted inferences (by adopting semantically rich formalisms as OWL 2). Finally,
allow managing more articulated specifications with detailed noteworthy is the support for an articulated discovery through
constraints. Moreover, the proposed approach fully supports both exact and approximated matches formally grounded on
proactive agents, which can fire periodic or sporadic internal service/resource composition.
events to trigger collaborative service discovery and MAS Quantitative performance results of the proposed approach
configuration updates. Finally, the small MAS described in are not provided here, but the semantic service discovery and
the example can be federated with other MASs in nearby orchestration based on Concept Covering is arguably the most
zones (e.g., of adjacent houses) by means of social interaction computationally demanding task, while social relationship
capabilities, ensuing from the possibility to establish friendship management is not resource-intensive. Results obtained in
or follower relationships between agents across zones. This [23] for an earlier version of this framework allow optimistic
allows taking advantage of sensing/acting capabilities of a expectations about feasibility on IoT device networks and
larger agent pool, as well as compensating possible deficits compatibility with performance requirements of HBA and AmI
of individual agents and zones reaching a concrete ambient scenarios.
intelligence in real-life significant scenarios. V. C ONCLUSION
B. Evaluation The paper proposed a novel semantic-based social MAS
framework. Though presented in a HBA scenario, features
In order to assess both peculiarities and capabilities of and approach are general-purpose and target several possible
the proposed semantic-based social MAS, a systematic com- Ambient Intelligence records. The application domains are
parison with existing IoT-oriented AmI approaches has been basically inherited from ontologies modeling the reference
carried out. Particularly, HBA platforms have been selected implementation.
as reference systems. In more detail, the following solutions The proposed approach enables autonomic agent interaction
have been considered: KNX IoT1 ; IzoT Platform2 , originally and a semantic-enhanced service/resource discovery grounded
developed by Echelon Corporation for the Industrial IoT but on the formal annotation of devices, environment and phe-
also exploited for HBA applications; Dog Gateway3 [20]; nomena. A case study and a comparison with state- of-the-art
Eclipse SmartHome4 . techniques help highlighting peculiarities of the proposal.
Table II highlights most relevant elements: it emerges that, Future work will include further investigation and extension
to the best of our knowledge, only the approach proposed of the social presence capabilities of agents, as well as novel
1 http://www.knx.org/knx-en/Landing-Pages/KNX-IoT
interaction patterns. A full prototypical implementation is
2 http://www.echelon.com/izot-platform expected to evidence possible optimization directions and scal-
3 http://dog-gateway.github.io/ ability concerns. Finally, graphical visualizations of devices’
4 http://www.eclipse.org/smarthome/index.html walls are being implemented.
101
ACKNOWLEDGEMENTS [20] D. Bonino, E. Castellina, and F. Corno, “The DOG gateway: enabling
ontology-based intelligent domotic environments,” IEEE Transactions
This work was supported in part by Italian PON project ERHA on Consumer Electronics, vol. 54, no. 4, pp. 1656–1664, 2008.
(Enhanced Radioteraphy with HAdrons). [21] L. Mainetti, V. Mighali, L. Patrono, and P. Rametta, “A novel rule-based
semantic architecture for iot building automation systems,” in Software,
R EFERENCES Telecommunications and Computer Networks (SoftCOM), 2015 23rd
International Conference on. IEEE, 2015, pp. 124–131.
[1] C. Savaglio, G. Fortino, M. Ganzha, M. Paprzycki, C. Bădică, and [22] S. N. Han, G. M. Lee, and N. Crespi, “Semantic context-aware service
M. Ivanović, “Agent-Based Computing in the Internet of Things: A composition for building automation system,” IEEE Transactions on
Survey,” in International Symposium on Intelligent and Distributed Industrial Informatics, vol. 10, no. 1, pp. 752–761, 2014.
Computing. Springer, 2017, pp. 307–320. [23] M. Ruta, F. Scioscia, G. Loseto, and E. Di Sciascio, “A semantic-enabled
[2] F. Sadri, “Ambient intelligence: A survey,” ACM Computing Surveys social network of devices for building automation,” IEEE Transactions
(CSUR), vol. 43, no. 4, p. 36, 2011. on Industrial Informatics, vol. 13, no. 6, pp. 3379–3388, dec 2017.
[3] W. Kastner, L. Krammer, and A. Fernbach, “State of the Art in [24] M. Ruta, F. Scioscia, G. Loseto, F. Gramegna, S. Ieva, A. Pinto, and
Smart Homes and Buildings,” in Industrial Communication Technology E. Di Sciascio, “Social Internet of Things for Domotics: a Knowledge-
Handbook, 2nd ed., R. Zurawski, Ed. CRC Press/Taylor & Francis, based Approach over LDP-CoAP,” Semantic Web, 2018, in press.
2014, ch. 55, pp. 1415–1434. [25] C. Bormann, A. P. Castellani, and Z. Shelby, “Coap: An application
[4] M. Ruta, F. Scioscia, and E. Di Sciascio, “Enabling the Semantic Web of protocol for billions of tiny internet nodes,” Internet Computing, IEEE,
Things: Framework and Architecture,” in Semantic Computing (ICSC), vol. 16, no. 2, pp. 62–67, 2012.
2012 IEEE Sixth International Conference on. IEEE, 2012, pp. 345– [26] M. Horridge and P. F. Patel-Schneider, “OWL 2 Web Ontology Language
347. Manchester Syntax (Second Edition),” W3C, W3C Working Group Note,
[5] L. Atzori, A. Iera, and G. Morabito, “From ”Smart Objects” to ”So- Dec. 2012, https://www.w3.org/TR/owl2-manchester-syntax/.
cial Objects”: The next evolutionary step of the Internet of Things,”
Communications Magazine, IEEE, vol. 52, no. 1, pp. 97–105, 2014.
[6] F. Scioscia, M. Ruta, G. Loseto, F. Gramegna, S. Ieva, A. Pinto, and
E. Di Sciascio, “Mini-ME matchmaker and reasoner for the Semantic
Web of Things,” in Innovations, Developments, and Applications of
Semantic Web and Information Systems. IGI Global, 2018, pp. 262–294.
[7] B. Parsia, S. Rudolph, M. Krötzsch, P. Patel-Schneider, and P. Hitzler,
“OWL 2 Web Ontology Language Primer (Second Edition),” W3C, W3C
Recommendation, Dec. 2012, http://www.w3.org/TR/owl2-primer.
[8] K.-Y. Lin and H.-P. Lu, “Why people use social networking sites: An
empirical study integrating network externalities and motivation theory,”
Computers in Human Behavior, vol. 27, no. 3, pp. 1152–1161, 2011.
[9] B. Ramis Ferrer, S. Iarovyi, L. Gonzalez, A. Lobov, and J. L. Mar-
tinez Lastra, “Management of distributed knowledge encapsulated in
embedded devices,” International Journal of Production Research, pp.
1–18, 2015.
[10] G. Biamino, “A semantic model for socially aware objects,” Advances
in Internet of Things, vol. 2012, 2012.
[11] O. Voutyras, P. Bourelos, S. Gogouvitis, D. Kyriazis, and T. Varvarigou,
“Social monitoring and social analysis in internet of things virtual
networks,” in Intelligence in Next Generation Networks (ICIN), 2015
18th International Conference on. IEEE, 2015, pp. 244–251.
[12] G. Capurso, M. Ruta, F. Scioscia, and E. Di Sciascio, “Object
(B)logging: Semantic Self-Description for Cyber-Physical Systems,” in
Proceedings of the Doctoral Consortium, Challenge, Industry Track,
Tutorials and Posters @ RuleML+RR 2017, ser. CEUR Workshop
Proceedings, July 2017, vol. 1875.
[13] R. De Virgilio, E. Di Sciascio, M. Ruta, F. Scioscia, and R. Torlone,
“Semantic-based RFID Data Management,” in Unique Radio Innovation
for the 21st Century: Building Scalable and Global RFID Networks.
Springer, 2011, pp. 111–142.
[14] J. Bleecker, “A Manifesto for Networked Objects Cohabiting with
Pigeons, Arphids and Aibos in the Internet of Things,” in Human-
Computer Interaction with Mobile Devices and Services, International
Conference on, 2006.
[15] S. Benford, A. Hazzard, A. Chamberlain, and L. Xu, “Augmenting
a guitar with its digital footprint,” in New Interfaces for Musical
Expression, 15th International Conference on, 2015.
[16] J. Miranda, N. Makitalo, J. Garcia-Alonso, J. Berrocal, T. Mikkonen,
C. Canal, and J. M. Murillo, “From the Internet of Things to the Internet
of People,” Internet Computing, IEEE, vol. 19, no. 2, pp. 40–47, 2015.
[17] V. Pejovic and M. Musolesi, “Anticipatory mobile computing: A survey
of the state of the art and research challenges,” ACM Computing Surveys
(CSUR), vol. 47, no. 3, p. 47, 2015.
[18] J. Ploennigs, B. Hensel, H. Dibowski, and K. Kabitzsch, “BASont –
A modular, adaptive building automation system ontology,” in IECON
2012-38th Annual Conference on IEEE Industrial Electronics Society.
IEEE, 2012, pp. 4827–4833.
[19] Y. Evchina, J. Puttonen, A. Dvoryanchikova, and J. L. M. Lastra,
“Context-aware knowledge-based middleware for selective information
delivery in data-intensive monitoring systems,” Engineering Applications
of Artificial Intelligence, vol. 43, pp. 111–126, 2015.
102