Semantic IoT Middleware-enabled Mobile Complex Event Processing for

Integrated Pest Management

Francesco Nocera, Tommaso di Noia, Marina Mongiello and Eugenio Di Sciascio

Department of Electrical & Information Engineering, Polytechnic University of Bari, Via Orabona, 4, Bari, Italy

Keywords:

Complex Event Processing, Cyber-physical Systems, Ontologies, IoT Middleware.

Abstract:

Agricultural domain presents challenges typical of the Cyber-Physical Systems ﬁeld and the hard-core of in-

formation technology industry of the new generation, such as Cloud computing and Internet of Things. In fact,

modern agricultural management strongly relies on many different sensing methodologies to provide accurate

information about crop, climate, and environmental conditions.

In this paper we propose an approach to model a mobile-driven and distributed Complex Event Processing so-

lution which is enabled by an IoT middleware. The proposed framework is robust with reference to contextual

and environmental changes also thanks to the exploitation of an ontological model.

1 INTRODUCTION AND

MOTIVATION

More than one third of the world’s labour force is em-

ployed in agriculture ﬁeld, which is one of the most

dangerous of all sectors and many agricultural work-

ers suffer occupational accidents and ill health. Fur-

thermore, informal employment, as a percentage of

non-agricultural employment, exceeds 50 per cent in

half of the countries with comparable data. In one-

third of countries, it affects over 60 per cent of work-

ers. (Horne et al., 2016)

As pointed out in the last Meridian Health Re-

port, presented by The European House-Ambrosetti

about the allocation of the basic items of public ex-

penditure grouped into ten macro categories (includ-

ing Environment, Health,...), one of the macro objec-

tives identiﬁed by the Prevention Plan is the identi-

ﬁcation of strategies for reducing or prevent poten-

tially harmful environmental exposures and improv-

ing health outcomes. The main focus is on admin-

istration’s methods for agrochemical inputs in order

to assure adequate plant nutrition and plant protection

through organic nutrient sources and Integrated Pest

Management (IPM), respectively.

IPM is a pest management strategy formally devel-

oped in the 1950s by entomologists and other re-

searchers in response to a widespread development in

agricultural settings of pesticide resistance in insects

The report is available at http://www.ambrosetti.eu/.

and mites, outbreaks of secondary and induced insect

and mite pests resulting from pesticide use, and trans-

fer and magniﬁcation of pesticides in the environment

(Flint and Van den Bosch, 2012). The UN’s Food

and Agriculture Organisation (FAO)

deﬁnes IPM as

“the careful consideration of all available pest con-

trol techniques and subsequent integration of appro-

priate measures that discourage the development of

pest populations and keep pesticides and other inter-

ventions to levels that are economically justiﬁed and

reduce or minimize risks to human health and the en-

vironment”.

IPM emphasizes the growth of a healthy crop with

the least possible disruption to agro-ecosystems and

encourages natural pest control mechanisms. Initially

focusing on biological control of insects and mites in

agricultural systems, over the last 70 years IPM has

assumed an ampliﬁed role, encompassing manage-

ment of diseases and weeds as well as insects, mites

and other arthropods in agricultural, horticultural, and

urban settings. IPM emphasizes selecting, integrat-

ing, and implementing complimentary pest manage-

ment tactics to maintain pests economically accept-

able while minimizing negative ecological and social

impacts of pest management activities (Flint, 2012).

As depicted in Figure 1 (left hand side), the ma-

jor challenge in the agricultural sector is identiﬁed as

no timely help and inadequate knowledge ﬂow in con-

nection with weather data at the required time.

http://www.fao.org/

610

Nocera, F., Noia, T., Mongiello, M. and Sciascio, E.

Semantic IoT Middleware-enabled Mobile Complex Event Processing for Integrated Pest Management.

DOI: 10.5220/0006369506380645

In Proceedings of the 7th International Conference on Cloud Computing and Services Science (CLOSER 2017), pages 610-617

ISBN: 978-989-758-243-1

Figure 1: Challenges and proposed solution.

Agricultural domain presents challenges typically en-

countered in the realm of Cyber-Physical Systems

(CPSs), such as: incomplete information, limited

sources of information, knowledge and infrastructure

deﬁciency, inadequate help and no timely support, ex-

ternal disturbances (weather), limited control author-

ity (fertilizers cannot make a plant mature arbitrarily

fast), etc.. Thanks to the spreading of sensors and to

the diffusion of low-cost miniaturized computational

resources, we are able to build objects that produce

data and interact with each other thus producing a

network of interconnected things, data, processes and

services. On the other hand, CPSs are transforming

the agriculture industry.

To solve these limitations and disadvantages, this pa-

per proposes a smart infrastructure designed to pro-

cess heterogeneous data sources input such as sen-

sors data, weather data, and the agricultural knowl-

edge collected into an ontology.

The proposed solution is based on a Complex Event

Processing (CEP) (Luckham, 2002) module partially

executed on mobile devices by introducing an Internet

of Things (IoT) middleware (Razzaque et al., 2015;

Fersi, 2015) that by using a programming language

that supports reﬂection mechanism (Mongiello et al.,

2016) serves as an interface between IoT compo-

nents and the ontological knowledge. The proposed

approach make possible: (i) a smoother and homo-

geneous communication among elements and a (ii)

dynamic, conﬁgurable and extensible infrastructure.

The identiﬁed solution and the relative instantiation

in agricultural domain aims at becoming an instru-

ment used for raising awareness about the use of treat-

ments. In this way the farmer can (i) have access to

all information related to the domain of interest at re-

quired time, (ii) create personalized defense plan, (iii)

receive alerts of changing weather conditions and (iv)

receive notiﬁcation and recommendations about his

treatment plans.

The remaining of the paper is organized as follows:

Section 2 illustrates background and related work

while Section 3 introduces the proposed approach. In

the following section we instantiate and validate the

approach in a real scenario. Conclusion and future

work close the paper.

2 BACKGROUND AND RELATED

WORK

There is no uniﬁed deﬁnition of a CPS. Generally,

CPSs are deﬁned as physical and engineered systems

whose operations are monitored, coordinated, con-

trolled and integrated by a computing and commu-

nication core. Internet transformed how humans in-

teract and communicate with one another, revolution-

ized how and where information is accessed.. Sim-

ilarly, CPS will transform how humans interact with

and control the physical world around us. CPSs per-

ceive the physical world, process the various data by

computers, and affect and change the physical world

(Hu et al., 2012). He JiFeng presented the concepts

of “3C”: Computation, Communication, and Con-

trol. Examples of CPS include medical devices and

systems, transportation vehicles and intelligent high-

ways, defense systems, aerospace systems, robotic

systems, process control, building and environmen-

tal control and smart spaces. CPS interact with the

physical world, and must operate dependably, safely,

securely, and efﬁciently and in real-time. CPS is an

emerging area, which cannot work efﬁciently with-

out proper software that handles both the data and the

business logic. The software development of CPS is a

critical issue because of its complexity in a large scale

pragmatic system. Furthermore, an object-oriented

approach (OOA) is often used to develop CPS soft-

ware, which needs some improvements according to

the features of CPS.

IoT Middleware and domain knowledge are the soul

of a CPS. One of the most important challenge related

to CPSs is knowledge collection and representation

in a usable and efﬁcient way as it is a cumbersome,

time consuming and expensive process. Sharing of

the same knowledge in different applications and its

reuse without or with little modiﬁcations in solving

separate problems is a critical factor in knowledge

aggregation and dissemination. Among knowledge

representation techniques, ontology formalisation can

be used to model real world in a consistent, formal,

manageable and reusable way (Jasper et al., 1999;

Stevens et al., 2004). Ontologies can be used as a tool

to handle several aspects of knowledge management

such as knowledge representation, knowledge shar-

ing and reuse, knowledge classiﬁcation, knowledge

evolution and knowledge search and retrieval (Alavi

et al., 2001).

One of the most cited deﬁnitions of ontology as “a

shared and common understanding of a domain that

can be communicated between people and across ap-

plication systems”(Gruber et al., 1993) leaves the

room for several interpretations of the term (Guarino,

Semantic IoT Middleware-enabled Mobile Complex Event Processing for Integrated Pest Management

611

1998).

Various ontology-driven applications are developed

in agricultural domain including knowledge systems

(Ahsan et al., 2014; Pereira et al., 2012; Xie et al.,

2007; Malik et al., 2015; Di Noia et al., 2016), expert

systems (Kang and Gao, 2013; Barshe and Chitre,

2012) and advisory systems (Chaudhary et al., 2015).

The most inﬂuential works are the following.

(Pereira et al., 2012) presented a Recommender Sys-

tem constructed as a Semantic Model that gives infor-

mation on planning better intercropping.

(Ahsan et al., 2014) presented a knowledge Model

based on Ontologies to study the process of knowl-

edge acquisition, best practices in the agricultural do-

main. The knowledge Base contained Wheat Crop

information.

(Malik et al., 2015) presented an ontology designed

and developed for the agricultural domain in which

various entities and relationships are represented. A

knowledge intensive system is developed based on a

small ontology.

(Barshe and Chitre, 2012) proposed an Agrosearch

system; the prototypic interface uses a relational ar-

chitecture for keyword maintenance while ontology

can be saved in the shape of URI.

(Kang and Gao, 2013) demonstrated the ontology ap-

plication in an agriculture information retrieval sys-

tem. The process of knowledge retrieval analyses

the characters of agricultural domain and the search

mechanism was designed with knowledge retrieval

ontology .

(Walisadeera et al., 2013) designed farmer centered

ontology. Many motivation scenarios are made into a

model with set of questions and ontology was devel-

oped for all stages of farming lifecycle.

(Chaudhary et al., 2015) presented an advisory system

for the cotton farmers in Gujarat using cotton ontol-

ogy, web services and Mobile Application Develop-

ment Advisory system.

In recent developments, monitoring and agriproduct

systems are developed using various others technolo-

gies. (Bo and Wang, 2011) presented and discussed

in their work: (i) advantages of IOT and Cloud Com-

puting in agriculture ﬁeld; (ii) various applications of

cloud computing and IOT in Agriculture and Forestry

on safety of Agriproduct, agricultural information

transformation and intelligent detection, precision ir-

rigation and forest identiﬁcation.

The IOT and Cloud Computing applications con-

tribute huge development towards agricultural sector

but it does not satisfy to meet all current challenges.

IOT is closely related to Cloud Computing in a way

that IOT obtains massive computing tools through

cloud computing and cloud computing ﬁnds the best

practicing channel based on IOT. Limited sources of

Information, the agricultural data is scattered in dif-

ferent places, domains making it difﬁcult to ﬁnd the

right information at the right time.

In the agricultural ﬁeld, ontologies may provide farm-

ers timely support to acquire information. Indeed,

there are many well-established and authoritative con-

trolled vocabularies, such as AGROVOC

, CAB The-

saurus

and NAL Thesaurus

AGROVOC is the most used vocabulary in several

systems, covering all areas of interest of the FAO, in-

cluding food, nutrition, agriculture, ﬁsheries, forestry,

environment etc. It is published by FAO and edited

by a community of experts. It is possible to extract

ontological concepts from AGROVOC thesaurus and

use them to build domain speciﬁc ontologies, retriev-

ing and organizing data in agricultural systems. These

practices are important in order to make this informa-

tion accessible to any user, especially regardless of

age.

3 PROPOSED APPROACH

We now illustrate the proposed approach and the

related CPS architecture identiﬁed. The whole in-

frastructure consists of two parts: the backend server

consisting an IoT middleware that automatically

perform actions according to the values received by

the sensors of devices by matching the set of formal

rules and the mobile device part principally based on

a CEP module. Figure 2 shows a graphical schema

of the proposed approach represented by means of an

abstract architecture.

For the backend server part(components con-

tained in the light blue box in Figure 2) we propose

the adoption of a reﬂective paradigm (Buschmann

et al., 1996) for modeling an IoT middleware, by

implementing the software design pattern Reﬂection.

The main concept in Reﬂection pattern is the dis-

tinction between base-level, meta-level and the

Meta-Object Protocol (MOP) (Buschmann et al.,

2007). The base-level contains the application logic

basic software while the meta-level takes care of all

the aspects that can change independently from the

base level.

A meta-level contains those parts of the application

that may vary at run-time, with the goal of creating

new applications starting from existing base-level

http://www.fao.org/agrovoc

http://www.cabi.org/

http://agclass.canr.msu.edu/

CLOSER 2017 - 7th International Conference on Cloud Computing and Services Science

612

Figure 2: Graphical schema of proposed model.

classes. The adaptation of the meta level is performed

indirectly thanks to the MOP, a speciﬁc Interface that

also makes possible the change of connections be-

tween the base-level and meta-objects. A meta-object

is deﬁned as an object that manipulates, creates,

describes, or implements other objects (including

itself)(Buschmann et al., 2007). In this way, by using

a programming language that supports reﬂection we

can design a completely conﬁgurable, extensible sys-

tem which adapts to different operating environments

where it is also possible to change the structure of

the objects themselves at run-time and then make the

software system more ﬂexible.

The message broker component models the Pub-

lish/Subscribe mechanism for deﬁning the rela-

tionships between messages containing physical

sensors information, and actions. The proposed

approach maps an IoT middleware on the three levels

of a reﬂection pattern according to the following

matching: the Condition to the Meta level, the Action

to the Base level and the Rule to the Meta-Object

Protocol. RULES are (i) modeled through a data

transformation module from the ontology and (ii)

stored in a RULE-BASED SYSTEM. A RULE ENGINE

is at the basis of the reasoning algorithm. The

Adaptor component works as a driver and translates

the received command in a real action. It is possible

to use place different Adaptor for each action.

The mobile device part(components enclosed in

the green box in Figure 2 ) includes ﬁve components.

The core component is the Complex Event Process-

ing (CEP) engine. In recent years, CEP over event

streams has become an increasingly important tool

to extract relevant situational knowledge from dis-

tributed systems in real-time or “near real-time”.

The concept of CEP was introduced by David Luck-

ham in his seminal work (Luckham, 2002) as a

“deﬁned set of tools and techniques for analyzing

and controlling the complex series of interrelated

events that drive modern distributed Information Sys-

tems(IS)”. Event pattern rules are used in CEP to

create complex events representing the combined ac-

tivities of sets of atomic events (Luckham, 2002).

The key characteristic of a CEP system is its ca-

pability to handle complex event situations, detect-

ing patterns, creating correlations, aggregating events

and making use of time windows. Speciﬁcation lan-

guages for event patterns are frequently inspired by

regular languages and therefore have automata based

semantics (Hopcroft et al., 2001). Several CEP sys-

tems have been developed in the last few years, each

one proposing a different processing model. Cur-

rently, the most popular are: StreamBase CEP

, Es-

per

, Apache Flink

, StreamDrill

. Sensor Adapters

connect to sensors and translate information obtained

from sensors into events format. All formatted events

are sent to Event Stream Management component,

which dispatches event streams to other components

such as the CEP engine or action handlers or sends the

events to the server through publish methods. Pattern

Deployment deploys/un-deploys the patterns, which

are dispatched by server to the CEP engine on the

mobile device. In order to interact with a user, Ac-

tion Handler execute actions like playing alarm audio,

displaying alerts and/or recommendations on smart-

phone in case the pre-deﬁned trigger events are de-

tected. The two part communicate through a pub-

lish/subscribe middleware composed by a Distributed

Service Bus (DSB), a Google Cloud Messaging ser-

vice and a Subscription Web Service.

https://www.streambase.com/

http://www.espertech.com/esper/

http://ﬂink.apache.org

https://streamdrill.com/

Semantic IoT Middleware-enabled Mobile Complex Event Processing for Integrated Pest Management

613

4 INSTANTIATION AND

VALIDATION OF THE MODEL

In this Section, to explain the abstract architecture de-

ﬁned in Section 3 let us consider its instantiation in a

real use case scenario. We instantiate the model for

an Italian company.

Figure 3: Architectural schema of server part.

For the mobile device part (see Figure 2), consid-

ering the limited computing resources on mobile de-

vices we use the light weight Esper engine as CEP

engine. Esper enables rapid development of appli-

cations that process large volumes of incoming mes-

sages or events, regardless of whether messages are

historical or real-time in nature. Esper ﬁlters and

analyzes events in various ways, and responds to

conditions of interest. Esper language speciﬁcation

is called Event Processing Language (EPL). It is a

SQL-like language with SELECT, FROM, WHERE,

GROUP BY, HAVING and ORDER BY clauses.

As depicted in Figure 3, we choose the following

technologies for the backend server part described in

Section 3 (Figure 2):

• DeviceHive

as IoT middleware. The reasons

for this choice are: ease of installation, the

rich documentation, the high integration with a

wide range of programming languages and IoT

protocols. In particular, we used the shell instance

http://devicehive.com/

Cloud Playground;

• Redis as a Message Broker. Redis

is a NoSQL

DBMS and allows a system to translate a message

from the messaging protocol of the sender to

the recipient’s messaging protocol. Through a

decoupling between publishers and subscribers,

Redis guarantees a greater scalability and a

dynamic network topology;

• Node JS as Event-driven component. It receives

data about the sensor and the variable of the

sensor through a REST interface;

• Observer. This component observes rules ex-

tracted from the Rule based-systems. In our im-

plementation, the rule based systems is a conﬁgu-

ration ﬁle.

The system will automatically perform actions ac-

cording to the values received by the sensors of de-

vices by matching the set of formal rules. Variables

to be monitored are air, temperature, humidity, ﬁne

dust emissions.

The hardware used to provide a source of data is

based on the Arduino platform, a small electronics

board equipped with a microcontroller, which is used

to quickly achieve hardware prototypes. In addition

to the Arduino board we have used the RaspberryPi

B+ board to send and receive data from/to the local

server which then makes a transition of data to Cloud

DeviceHive servers.

Within the component, a conﬁguration JSON ﬁle con-

tains the deﬁnition of the rules on ontology data ex-

tracted and the actions to take if the rule is ﬁred. The

OBSERVER (of the data ﬂow) notiﬁes the NODE JS

that forwards to the MESSAGE BROKER about the ex-

tracted active rule. The active rule together with the

information about sensors and variables, (the predi-

cate of our Condition) is published on the Redis chan-

nel. Information about the arrival of new data is pub-

lished on the message channel . The MESSAGE BRO-

KER works as a through for data ﬂow and the active

rule that are forwarded to the Reﬂective part. Re-

ceivers subscribe to the MESSAGE BROKER and are

notiﬁed of the message. The MESSAGE BROKER for-

wards the whole message made up of sensor, variable,

and the rule, (extracted from the conﬁguration ﬁle) to

the Rule Engine that executes it.

Reﬂection is applied thanks to the signing of the re-

ceiver component of the Redis channel.

The Condition Level and the RULE ENGINE enable

http://redis.io/

CLOSER 2017 - 7th International Conference on Cloud Computing and Services Science

614

the action level (Figure 3 depicts the actions imple-

mented) of the reﬂective component.

In this instantiation we constructed an OWL 2 on-

tology called Agrimag based on the AGROVOC the-

saurus. The ontology contains the main knowledge

related to the agricultural ﬁeld, in particular the as-

pects concerning pest management and control, with

a view to respecting the environment and operator

health. Figure 4 shows a screenshot of the Prot

editor GUI illustrating the Class and Object Property

hierarchies of our ontology.

The ontology describes the main concepts related

to the management of cultivation and chemical treat-

ments necessary to ensure agricultural production,

with particular reference to the guidelines issued by

the Italian Ministry of Agriculture.

Table 1: Ontology metrics.

Metrics

Axiom 7694

Logical axiom count 6297

Class count 32

Object property count 35

Data property count 07

Individual count 2012

Class axioms

SubClassOf axioms count 24

DisjointClasses axioms count 02

Object property axioms

SubObjectPropertyOf 18

InverseObjectProperties 03

ObjectPropertyDomain axioms count 21

ObjectPropertyRange axioms count 22

Data property axioms

SubDataPropertyOf 01

FunctionalDataProperty 01

DataPropertyDomain 05

DataPropertyRange 05

Individual axioms

ClassAssertion axioms count 1893

ObjectPropertyAssertion axioms count 3161

DataPropertyAssertion axioms count 1139

Annotation axioms axioms

AnnotationAssertion 1323

The most important classes are the following:

< Crop >: identiﬁes the planted crop in the ﬁeld.

Each culture is characterized by a speciﬁc set of

harmful agents that affect, as well as by a set of active

ingredients and commercial authorized products;

< Pest >: modeling agents harmful to plants (eg.

Mushrooms, Insects, Weeds,etc.), which the farmer

has to worry about ﬁnding a remedy through the use

http://protege.stanford.edu/

of Active Substance;

< ActiveSubstance >: is a chemical compound used

in agriculture as an antagonist of one or more kinds

of pathogens/pests;

< CommercialProduct >: indicates a formulation

commercially available. Its spectrum of action is

determined by its composition (made of ActiveSub-

stance), as well as legislative constraints, regulations

and authorizations imposed by the Ministry of

Health;

< InstitutionalActiveSubstanceU sage >: models

a ternary relationship that relates a Pest with an

authorized ActiveSubstance to a speciﬁc crop.

Table 1 summarizes the metrics associated to our

ontology which is publicly available.

Possible use

cases identiﬁed for a farmer are the following:

• query the ontology via a mobile app or company

Web page, to learn about the solutions to be im-

plemented (interventions and curative active sub-

stance, notes) to cope with all possible diseases

for each crop;

• the mobile device through the GPS module lo-

cates the exact position of the ﬁeld on the map and

can receive information related to defense plans

for their treated cultures;

• based on treated cultures (stored into mobile

device), it is possible to receive recommenda-

tions/alerts about their treatment (period, healing

product, doses, etc.)

Figure 5 shows a screenshot of the Web applica-

tion concerning the creation of the defense plan. Once

selected the crop (Cherry tree) and one related disease

(Archips rosanus), the farmer mobile device displays

the information solution for each selected disease.

5 CONCLUSION AND FUTURE

WORK

Agriculture in urban areas have become a new trend.

In Cyber-Physical Systems(CPSs) perspective, this

domain presents typical challenges such as incom-

plete information, limited sources of information,

knowledge and infrastructure deﬁciency, inadequate

help and no timely support, limited control authority

and so on. One of the challenges in building CPS is

the way software will be developed and composed on

the top of ﬂexible infrastructures and integration ar-

chitectures.

A full description of Ontology metrics is available at

http://protegewiki.stanford.edu/

Semantic IoT Middleware-enabled Mobile Complex Event Processing for Integrated Pest Management

615

Figure 4: Entities Tab Prot

Figure 5: Mobile app screenshots.

To solve the limitation, disadvantages and challenges

described, in this paper we present an infrastructures

for distributed CEP that will be partially executed on

mobile devices by introducing server side a Reﬂec-

tive IoT Middleware. The main advantages of the

system are: (i) using CEP engine at system run-time

enables event-driven monitoring and update notiﬁca-

tions, and (ii) modeling system with human-machine

understandable ontologies ensures easier reconﬁgura-

tion of the system. To validate the model we propose

an instantiation in a real scenario also designing an

OWL 2 Ontology that encodes knowledge about as-

pects related to Integrated Pest Management practice.

The solution identiﬁed aims at becoming a useful in-

strument by ecosystem actors for raising awareness

about the use of chemical treatments.

In this stage of the work, we performed a ﬁrst set

of experiments to validate the approach providing the

tool to a Company and testing single components. We

are current working to extend the system for the cre-

ation of an Advanced Cyber-Physical System.

ACKNOWLEDGEMENTS

The authors acknowledge support of Simone Salerno

for ontology building. Francesco Nocera acknowl-

edges support of Exprivia S.p.A Ph.D grant 2016.

CLOSER 2017 - 7th International Conference on Cloud Computing and Services Science

616

REFERENCES

Ahsan, M., Motla, Y. H., and Asim, M. (2014). Knowledge

modeling fore-agriculture using ontology. In Open

Source Systems and Technologies (ICOSST), 2014 In-

ternational Conference on, pages 112–122. IEEE.

Alavi, M., Leidner, D. E., et al. (2001). Knowledge manage-

ment and knowledge management systems: Concep-

tual foundations and research issues. MIS quarterly,

25(1):107–136.

Barshe, P. and Chitre, P. (2012). Agricuture system based

on ontologyagrosearch. Int. J. Emerg. Technol. Adv.

Eng, 2(8).

Bo, Y. and Wang, H. (2011). The application of cloud com-

puting and the internet of things in agriculture and

forestry. In Service Sciences (IJCSS), 2011 Interna-

tional Joint Conference on, pages 168–172. IEEE.

Buschmann, F., Henney, K., and Schmidt, D. C. (2007).

Pattern-Oriented Software Architecture, Volume 4, A

Pattern Language for Distributed Computing. Wiley.

Buschmann, F., Meunier, R., Rohnert, H., Sommerlad, P.,

and Stal, M. (1996). Pattern-oriented software archi-

tecture: a system of patterns. John Wiley & Sons, Inc.,

New York, NY, USA.

Chaudhary, S., Bhise, M., Banerjee, A., Goyal, A., and

Moradiya, C. (2015). Agro advisory system for

cotton crop. In 2015 7th International Conference

on Communication Systems and Networks (COM-

SNETS), pages 1–6. IEEE.

Di Noia, T., Marina, M., Francesco, N., and Eugenio, D. S.

(2016). Ontology-based reﬂective iot middleware-

enabled agriculture decision support system.

Fersi, G. (2015). Middleware for internet of things: A

study. In Distributed Computing in Sensor Systems

(DCOSS), 2015 International Conference on, pages

230–235. IEEE.

Flint, M. L. (2012). IPM in practice: principles and meth-

ods of integrated pest management, volume 3418.

UCANR Publications.

Flint, M. L. and Van den Bosch, R. (2012). Introduction

to integrated pest management. Springer Science &

Business Media.

Gruber, T. R. et al. (1993). A translation approach to

portable ontology speciﬁcations. Knowledge acqui-

sition, 5(2):199–220.

Guarino, N. (1998). Formal ontology and information sys-

tems. 98(1998):81–97.

Hopcroft, J. E., Motwani, R., and Ullman, J. D. (2001). In-

troduction to automata theory, languages, and compu-

tation. ACM SIGACT News, 32(1):60–65.

Horne, R., Khatiwada, S., and Kuhn, S. (2016). World em-

ployment and social outlook: trends 2016.

Hu, L., Xie, N., Kuang, Z., and Zhao, K. (2012).

Review of cyber-physical system architecture. In

Object/Component/Service-Oriented Real-Time Dis-

tributed Computing Workshops (ISORCW), 2012 15th

IEEE International Symposium on, pages 25–30.

Jasper, R., Uschold, M., et al. (1999). A framework for un-

derstanding and classifying ontology applications. In

Proceedings 12th Int. Workshop on Knowledge Acqui-

sition, Modelling, and Management KAW, volume 99,

pages 16–21.

Kang, J. C. and Gao, J. L. (2013). Application of ontol-

ogy technology in agricultural information retrieval.

In Advanced Materials Research, volume 756, pages

1249–1253. Trans Tech Publ.

Luckham, D. (2002). The power of events, volume 204.

Addison-Wesley Reading.

Malik, N., Sharan, A., and Hijam, D. (2015). Ontol-

ogy development for agriculture domain. In Com-

puting for Sustainable Global Development (INDIA-

Com), 2015 2nd International Conference on, pages

738–742. IEEE.

Mongiello, M., di Noia, T., Nocera, F., di Sciascio, E., and

Parchitelli, A. (2016). Context-aware design of re-

ﬂective middleware in the internet of everything. In

Federation of International Conferences on Software

Technologies: Applications and Foundations, pages

423–435. Springer.

Pereira, D. H. G., Dantas, C. F. F., and Ribeiro, C. M.

(2012). A pragmatic approach for sustainable devel-

opment based on semantic web services. In Proceed-

ings of the 14th International Conference on Informa-

tion Integration and Web-based Applications & Ser-

vices, pages 82–90. ACM.

Razzaque, M., Milojevic-Jevric, M., Palade, A., and Clarke,

S. (2015). Middleware for internet of things: a survey.

Internet of Things Journal, IEEE, PP(99).

Stevens, R., Wroe, C., Lord, P., and Goble, C. (2004). On-

tologies in bioinformatics. In Handbook on ontolo-

gies, pages 635–657. Springer.

Walisadeera, A. I., Wikramanayake, G. N., and Ginige, A.

(2013). Designing a farmer centred ontology for so-

cial life network. In DATA, pages 238–247.

Xie, N., Wang, W., and Yang, Y. (2007). Ontology-based

agricultural knowledge acquisition and application.

In International Conference on Computer and Com-

puting Technologies in Agriculture, pages 349–357.

Springer.

Semantic IoT Middleware-enabled Mobile Complex Event Processing for Integrated Pest Management

617