2018 IEEE 6th International Conference on Future Internet of Things and Cloud

A Self-organized Framework for Insurance Based on Internet of Things and

Blockchain

Monireh Vahdati Kamran Gholizadeh HamlAbadi

Faculty of Computer and Information Technology Faculty of Computer and Information Technology
Engineering Engineering
Qazvin Branch, Islamic Azad University Qazvin Branch, Islamic Azad University
Qazvin, Iran Qazvin, Iran
[email protected] [email protected]

Ali Mohammad Saghiri Hassan Rashidi

Computer Engineering and Information Technology Associate Professor, Department of Mathematics and
Department Computer Science
AmirKabir University of Technology Allameh Tabataba'i University
Hafez Ave., Tehran, 15914, Iran Tehran, Iran
[email protected] [email protected]

Abstract— Recently, Internet of Things (IoT) and blockchain companies have few devises to characterize the users
technology are used to improve the traditional companies. One considering their risks. in other word, a low-risk user is used
type of traditional companies is insurance companies. These for Compensation. Another drawback of these companies is
companies suffer several problems that are leading to decrease that the existing algorithms are not transparent. many of
their efficiency in modern societies. Many users of these insurance companies do not like to publish the documents of
companies are unhappy and dissatisfied because of some their turnover. In addition, many of traditional algorithms are
problems. The main problems of these companies are that their not online. the changes in the decision makings in the
operations are not transparent. Moreover, their insurance
company are very slow. this means that their algorithms are
rates are not calculated and in adaptive with customer’s
characteristics, e.g for low-risk and high-risk customers. Note
not adaptive. In recent years, modern technologies are used
that, users are not identical to the same insurance rates, which to solve some of the mentioned problems. One of these
are determined by the insurer. In modern society, we have the technologies is peer-to-peer insurance In 2010, peer-to-peer
IoT technology that enables us to find some specific insurance was first introduced in the German insurance
information about users. This information enables the market [1]. In 2013, about many of those who took
insurance companies to increase its accuracy. In addition, advantage of the peer-to-peer-insurance model were repaid
considering the potential of the smart contract of the contributions [2]. In 2014, the a British insurance company
blockchain, determining insurance rates must be transparent. known as Guevara introduced the peer-to-peer insurance
This article proposes a self-organizing framework for concept for car insurance in [3]. A company called Riovic
insurance based on IoT and blockchain, which eliminates the was launched in South Africa [4]. Company insPeer was
main problems of traditional insurance. To evaluate the launched in France. In 2016, Huddle Money was launched in
proposed framework, it is compared with the existing Australia, initially with a peer-to-peer model of insurance.
framework. The results show that it has several advantages In other hand, we have the blockchain technology and
over the existing framework and could be useful in near future. Internet of Things (IoT) [5][6][7]. These technologies are
able to solve many of problems which are mentioned before.
Keywords- Internet of Things (IoT); Blockchain; Insurance; In modern society, we have the following capabilities.
Cognitive System
x the IoT technology that enables us to find some
I. INTRODUCTION specific information about users. This
Recently, many of traditional companies such as information enables the insurance companies to
insurance companies have been evolved. Traditional increase its accuracy. In addition, considering
algorithm for managing insurance companies has many the potential of the smart contract of the
Weaknesses which some of them are described as follow. blockchain, determining insurance rates must be
These algorithms are not fair. In a traditional insurance transparent[8].
companies, both high-risk and low-risk users may have x The cognitive systems that enable us to
considered as equal users. This is because the traditional understand users based on past communication

978-1-5386-7503-8/18/$31.00 ©2018 IEEE 169

DOI 10.1109/FiCloud.2018.00032
 and behavior. To do that, these systems utilize Smart contracts enforce the malicious party’s payments, and
evidence-based reasoning. recently, cognitive therefore miners need neither trust the attacker’s intentions
systems help insurers offer engaging and nor his ability to pay. Assuming pool members are rational,
personalized interactions with users. In the near an adversary with a single mining ASIC can, in theory,
future, cognitive computing could predict the destroy all big mining pools without losing any money (and
risks of users in every fields. These systems even make some profit) [15].
may use to suggest the final policy in the The objective of the work reported in [16] is to support
insurance company[9]. actors involved in this decision making process by
There is one work reported in [10] in which there fields: illustrating the capabilities of the blockchain technology.
cognitive systems, blockchain, and IoT are combined which The approach reported in [17] purpose of such system
each other to bring a solution for insurance companies. In rises as result of an evaluation for the payment systems of
other hand, no general framework is reported in the the vehicle insurance based on the method we drive the car.
literature to utilize the mentioned fields simultaneously. The In this approach, Arduino, Proteus simulation environment
study and implementation in [10] is very dependent on IBM for verifying the circuit functionality, and a smart phone are
Watson and other products of this company. Another used.
drawback of this work is that it does not support fully . In this paper, we suggested a framework for self-
distributed computation required by cryptocurrencies such organized insurance based on cognitive systems,
as bitcoin. blockchain, and IoT. The suggested framework can be used
This article proposes a self-organizing framework for as a standard framework for the next generation of
insurance based on IoT and blockchain, which eliminates the insurance companies. note that, a self-organized framework
main problems of traditional insurance companies. To for insurance companies based on IoT, blockchain
evaluate the proposed framework, it is compared with the technology, and cognitive systems has not been reported.
existing framework. The results show that it has several
advantages over the existing framework and could be useful III. THE PROPOSED MODEL
in near future. In this section, we present a self-organized framework for
The rest of the paper is organized as follows. Section II insurance based on IoT and blockchain Technology. We
summarize the related work. In Section III, the proposed used a cognitive systems introduced by [18][19][20][9]. The
model is described. Section IV proposed a case study; structure of the proposed framework is given in Fig. 1. This
Section V implementation is described. In section VI framework consists of three layers: Requirement Layer,
Discussion and Comparison with other studies, Finally, Cognitive Process Layer, and Adaptive Insurance Layer. In
conclusion and future research are drawn in Section V. the Adaptive Insurance Layer, the elements of insurance are
combined with blockchain technology and IoT. Note that, the
II. RELATED WORKS combination between IoT and blockchain technology creates
Recent approaches for designing modern companies many management problems which should be solved by
focus on IoT and Blockchain technology. utilizing IoT in cognitive systems. In the proposed framework, the cognitive
insurance domain has some challenges. One of them focus process layer is in charge of managing the system by solving
on security aspect of IoT it means the risk of connecting to management problems of the Adaptive Insurance Layer.
internet. Another one is privacy and security information of The details of each layer are given in the rest of this section.
customers shared on the internet. Cyber-attack is one of the
tremendous problem treats many insurance businesses. Cognitive Specification Language Requirement Layer
Also, malfunction smart devices transfer wrong information
to other devices and it causes incorrect data to the insurer Goal

[11].
Designing the distributed platform in order to support
transaction execution in insurance process by using Cognitive Engine Cognitive Process Layer

blockchain as a system service is a key idea reported in [12].

In this paper, there are some concerns which can be consider Act

such as: transaction processing time, payment settlement Observe

time and security protection of process.

Blockchain as a breakthrough technology has lots of Blockchain Unit Peer to Peer Communication Manager

Adaptable Insurance Layer

benefits that can be represented in insurance company
Cryptocurrency Manager
however should consider company’s sections and goals. Smart Contract

This technology has its pros and cons which are reported in
[13]. Figure 1. A Self-Organized framework for Insurance
Utilizing the blockchain concept in order to share fraud
intelligence data in insurance system is reported in [14].

170
A. Requirement Layer cognitive process layer and other units of the
In the requirement layer, a language called Cognitive Adaptive Insurance Layer in the system. This
Specification Language (CSL) is used to determine the goal unit also manages the information about the
and behavior of the network. This language is used to write wallets of the users. It should be noted that the
on a file called configuration file. This file is shared among type of coins of the users should be managed in
those entities which manage the configurations of the this unit. In other word, the payment manager
Insurance system. It should be noted that, changing the unit must be able to communicate with the
goals in the requirements layer leading to changing the platform of every type of coins. All
optimizing functions in the cognitive process layer. The cryptocurrencies in Table 1 have been used in
goals of the systems are determined by some commands insurance.
obtained from the voice (or speech), command line, or any
TABLE I. CRYPTOCURRENCY IN INSURANCE
type of direct interaction between users and the system.
Some features of the goals which can be determined by the Cryptocur Description Reference
following elements. rency in
x Insurance Type Insurance
Bitcoin This cryptocurrency is the first prototype of
x Cryptocurrency Type cryptocurrency and is invented to enable
[16][21]
x Smart Contract Type money transfers among parties without relying
x Sensor Type on intermediaries.
Fedcoin ‘Fedcoin,’ dates back to the original design
x Actuators Type [21]
goal of Bitcoin for insurance purpose
The above elements can be used in commands. The Umbrella This cryptocurrency is based on Ethereum
Coin blockchain. This cryptocurrency benefits for
commands can be also used to tune the system. The hidden costs of insurance. They plan for the [22]
commands can be fetched from outside of the framework. policy/claim system to be completely
Considering distributed nature of IoT for sharing the democratized.
configuration things the blockchain can be used. Ethereum This cryptocurrency is used to provide
insurance benefits for hidden costs of [16]
insurance
B. Adaptive Insurance Layer
Bitpark Bitpark is a new insurance service with P2P
This layer provides required information for the cognitive Coin Insurance and Cryptocurrency.
[23]
process layer and then operates on manageable elements of Etherisc In this cryptocurrency, we are building a free,
the systems. The detailed information of this layer is given open-source, open-access platform for [24]
decentralized insurance.
in the rest of this section. This layer contains four units
described as bellow.
x Blockchain unit: this unit in charge of C. Cognitive Process Layer
managing required information in one or In this layer, the cognitive engine observes the
multiple blockchains. In this unit, every type of information about insurance and then executes appropriate
blockchain may be used. This unit has three algorithms for managing the system. This layer considers
sub-unit blockchain of things, blockchain of the goals of the self-organized insurance that are specified
microservices and blockchain of smart in requirement layer. In this layer several types of engines
contracts. For example, the information of may be designated which are described as bellow.
smart contracts used by the system are indexed x An engine for finding the goals of the insurance
in the blockchain of smart contracts. company by interpreting the configuration file. In
x Peer-to-Peer Communication Manager: This this engine, machine translation algorithms may be
element facilitates the thing to thing applied to extract the goals. Watson [25] can be
communication and data sharing using peer-to- used in this unit.
peer networking technology. x An engine for managing the complexities related to
x Smart contract: this element provides the smart contracts.
required functions for using the smart contracts x An engine for managing the complexities related to
defined in the system. Note that, the codes of payment process.
smart contracts are stored by the blockchain x An engine for managing peer-to-peer
unit in a blockchain [18]. The cognitive engine communication and blockchain.
is in charge of tuning and using the smart x An engine for predicting the characteristics of the
contract unit considering the goal of the system. users.
All contract related to insurance company are x An engine for predicting the insurance rate.
managed by smart contract unit. x An engine for generating appropriate insurance
x Cryptocurrency Manager: this unit supports policies.
the payment process in cooperation with

171
 Because of distributed nature of IoT, the cognitive x Vehicle whose information is first recorded in
engine can be implemented using one of the following the system and its user has a suitable history.
methods.
x Vehicle whose information is first recorded in
x Centralized approach: In this approach, the the system and its user does not have a suitable
cognitive engine is implemented in one server. record.
x Semi-centralized approach: In this approach, the
engines of the cognitive engine are implemented in x Vehicle information that is already in the system
multiple servers. and its user has a suitable history.
x Fully-distributed approach: In this approach, x Vehicle information that is already in the system
every thing has its own cognitive engine. and its user does not have a suitable history.
IV. A CASE STUDY: VEHICLE SMART INSURANCE According to one of scenarios in above and the two
In this paper, we will present a case study for vehicle scenarios in below, the insurance package is considered and
smart insurance. In this case study, each user and vehicle pricing for Insurance is done.
have an ID identifier and their information is saved in x Choosing smart insurance package and pricing
system. The system goal is determined in the Requirement it based on vehicle model and the distance
Layer. The goal is to select the type of insurance package for traveled by the car.
the customer. The goal can be set automatically by the x Pricing based on user behavior records.
system or manually by the user. The goal is entered into the
Cognitive Process layer and is received by the Cognitive
Engine. At this point, the vehicle information and user V. IMPLEMENTATION
behavior information in Table 2 are received by the third In this section, we give the implement details of the
layer and stored in the Cognitive Engine. These factors are proposed framework. At first, some basic concepts are
weighed and analyzed on the basis of risk metrics and then given, then a mapping used in the implementation domain is
used to select the insurance package. There are several described, and finally the mechanisms and related
insurance companies in one country in our system. But the algorithms are described.
system is integrated and obtains car insurance records from
the Rest APIs of their responsive insurance companies. In A. Basic Concepts
order to verify more precisely the user's behavior, user We introduce some basic concepts in the three next sub-
driving violations, accident, etc., it is obtained through the sections:
Rest API of the Police Department. This information is
analyzed along with vehicle information and user behavioral 1) Smart Contract
information in the Cognitive Engine and based on this smart A smart contract is a program intended to digitally manage
contract behavior are provided for the user. Through peer-to- the negotiation or performance of a contract. The execution
peer Insurance and cryptocurrency, we try to provide an
of a smart contract can be simple or very complicated.
insurance service that is both transparent and user-directed.
We will provide a variety of services, such as smart contract There are two types for execution of the smart contracts
technology. Smart contracts are used to ensure success of described as bellow[26]:
contractual duties, separation of billing management costs,
an approval and compensation system. x Installed: in this type, a business logic is used to
design an explicit smart contract.
TABLE II. BEHAVIORAL INFORMATION RECEIVED FROM SENSORS OF x On-chain: in this type, smart contract is executed
VEHICLE/USER
within a chain code. Source Code of this type pf
Order Obtained data through sensors contract is defined based on business logic.
1 Driving at night Considering the above type for the smart contracts, there
2 Driving on busy roads are three technologies for implementing smart contracts
3 Driving on high-risk roads which are described as bellow [26]:
4 Driving more hours on a day that has more traffic
5 Driving seasonally is the year that travel time
6 Weather conditions x Smart contract application engine: in this
7 Drowsiness of the driver technology, the engine is designated as library
8 A diagnostic test driver uses of drugs and alcohol functions.
x Chaincode: in this technology, the smart contact is
In some vehicles, the sensors in Table 2 are embedded in a program called chaincode. There are two of
the vehicle. However, sensors can be inserted through the
different type of chaincodes which are described as
Arduino and raspberry pi boards in vehicle and the existing
data can be entered into the system [17]. To provide an bellow.
intelligent insurance and user behavior analysis, we consider
four scenarios.

172
 x Transaction Families: in this technology, According to Fig. 2, in the proposed framework the
businesses can pick the level of versatility and risk traditional contract becomes the smart contract. The vehicle
that’s right for their network. is mapped to vehicle sensors. Further, cash money will
become Insurance cryptocurrency (INS-COIN). INS-COIN
2) Hyperledger is a coin based in insurance. INS-COIN aims at improving
Hyperledger is supported by the Linux Foundation, insurance care worldwide and making it affordable through
which can be used in Finance, Banking, IoT, SCM (Supply crowd power. The future step, user agent can use
Chain Management), production and modern technologies. credit/debit card to purchase INS-COIN directly on the
It is including reference architectures like Hyperledger insurance website. We can be used the stellar network for
Burrow, Hyperledger Fabric, Hyperledger Iroha and creating an INS-COIN Cryptocurrency. Using INS-COIN
Hyperledger Sawtooth [27]. Since, we selected the has advantages for the user agent. Any user agent that uses
Hyperledger Sawtooth in our implementation, its benefits INS-COIN cryptocurrency for its own payment can include
are listed below. variable discount. In the proposed framework, user behavior
is modeled in the cognitive engine. If the user agent is
x On-chain smart contract type. behaving appropriately, the above system will reward it with
x Transaction families smart contract technologies. INS-COIN and will charge the user's INS-COIN.
x Support language writing such as C++, Go, Java, Obviously, with a history of bad user agent behavior
JavaScript, Python, Rust, or Solidity (through Seth) insurance penalty is calculated based on INS-COIN. The
agents information has been stored in each blockchain.
3) Stellar Network
This network is a distributed payments infrastructure that C. Mechanism
connects banks, and payment systems. It provides the
The self-organized mechanism used for managing the
creating of general wallets, banking tools, smart contracts,
insurance company is explained as bellow.
and cryptocurrencies[28]. It uses its own form of consensus
called Stellar Consensus Protocol (SCP). This network has
Submission Phase: Customer agent submit information to
the following properties [28]:
insurance agent.
x Low latency.
x Digital signatures and hash families. Registration Phase: Insurance agent registers the request of
x Decentralized control. the customer.
x Flexible trust
Data Gathering Phase: If the request of the customer agent
B. Mapping is submitting for the first time then Goto Phase A Else Goto
Phase B.
For implementing a self-organized insurance, we replaced
traditional interaction by blockchain interaction. In our
Phase A: extract information from police agent
framework, we used agents, smart contract, vehicle sensors
Blockchain
and cryptocurrencies. Fig 2. Showed the mapping used in
Phase B: customer agent has history, the information
our framework.
gathers from different locations
x Sensors (on-board and out of vehicle which
contains: nearby vehicles, traffic light,
Traditional Contract Smart Contract
government authority monitors) gather
information Insurance agent Blockchain
x Police agent Blockchain

Vehicle Vehicle Sensors Analyzing Phase: Insurance agent analyzes the gather data
to find user’s behavior patterns and updates insurance rated
values
Insurance
Cryptocurrency
C h
Cash
(INS-COIN)
Recommendation Phase: Insurance agent sends
information to customer Blockchain, insurance Blockchain,
police Blockchain in an interface.
Vehicle Driver Police Officer Police Agent
User Agent

Insurance Company
Insurance Agent
Contract Execution Phase: Smart contract runs in a
blockchain based on Hyperledger Fabric and update all
Figure 2. The mapping used in our framework
blockchain such as insurance, customer and police

173
Payment Phase: Payment request is executed on the 2017 )
blockchain of cryptocurrency. At the end of this phase the [32] Not Reported 9 9
(Schwartz, 2016)
payment process is verified automatically. 9 9 9
[10]
(Patra, 2017)
Our Study 9 9 9

VII. CONCLUSSION AND FUTURE RESEARCH

In this paper, we proposed a self-organized framework
for Insurance based on Internet of Things, cognitive
systems, and blockchain technology. In this framework, the
cognitive engine of the cognitive process layer observes the
status of things from the Adaptive Insurance Layer. Then,
the cognitive engine acts on actuators of the things and
triggers an appropriate list of the smart contracts. In the
proposed framework, various coins (cryptocurrencies) can
be used for the payment process using peer-to-peer
communication protocols. Also, we presented a smart
insurance for vehicles. The proposed framework was
Figure 3. Self-Organized Mechanism compared with other existing solutions with respect to three
features. To the best of our knowledge, our work is novel
D. Implementation details and there hasn’t been any prior study on designing a self-
We implement the idea of the self-organized framework organized Insurance based on Internet of Things, cognitive
in insurance with Hyperledger. We used Hyperledger systems, and blockchain technology. The proposed
Sawtooth project for implement the proposed framework. In framework can be merged with the existing
addition, we used cognitive systems and on-chain smart cryptocurrencies of peer-to-peer insurance because all
contract, the Sawtooth has enough features and suitable for elements of the peer-to-peer insurance are embedded in our
our work. We can implement and maintain a the proposed framework [21] [23]. For future research we can apply the
framework using Docker [28]. We also used the stellar technologies of Web of Things in the proposed framework.
network for defining and managing INS-COIN. ACKNOWLEDGMENT
VI. DICUSSION AND COMPARISON Authors thanks to Mr. Hamid Reza Soltani, software
Table 3 compares the suggested algorithm with other expert in Sadad Informatics Corporation (SIC) in the Tehran-
Iran.
studies reported in the literature. Table 3 shows that most
related works focus on one of these three features. To the REFERENCES
best of our knowledge, our work is novel (unique) and there [1] M. Butcher, “Innovative Crowdsource Insurance Startup
hasn’t been any prior study on designing an algorithm based Friendsurance Raises New Cash,” 2014. [Online]. Available:
on internet of things, cognitive systems and blockchain https://techcrunch.com/2014/04/03/innovative-crowd-sourced-
technology for Insurance companies. insurance-startup-friendsurance-raises-new-cash/. [Accessed: 13-Apr-
2018].
[2] F. Aktuelle Meldungen, “Friendsurance zahlt sich aus,” 2014.
TABLE III. COMPARISON WITH OTHER INSURANCE SOLUTIONS
[Online]. Available: https://www.mittelstand-
nachrichten.de/finanzen/friendsurance-zahlt-sich-aus/. [Accessed: 13-
Reference Blockchain IoT Based Cognitive Apr-2018].
based Based
[3] B. SCHILLER, “Insuring Yourself With A Group Of Friends Has
[11] Not Reported 9 Not Reported
Benefits,” 2014. [Online]. Available:
(Jai, 2015)
https://www.fastcompany.com/3032809/insuring-yourself-with-a-
[12] 9 Not Reported Not Reported
group-of-friends-has-benefits. [Accessed: 13-Apr-2018].
(Mazumdar, 2018)
9
[4] M. Sikupela, “Riovic wants to be the personal financial advisor in
[14] Not Reported Not Reported
(Nath, 2016) your pocket,” 2015. [Online]. Available:
[16] 9 Not Reported Not Reported http://ventureburn.com/2015/12/riovic-wants-to-be-the-personal-
(Gatteschi, 2018) financial-advisor-in-your-pocket/. [Accessed: 13-Apr-2018].
[29] Not Reported 9 Not Reported [5] K. Christidis and M. Devetsikiotis, “Blockchains and Smart Contracts
(Chapman, 2015) for the Internet of Things,” IEEE Access, vol. 4, pp. 2292–2303,
[30] Not Reported Not Reported 9 2016.
(Forster, 2017) [6] M. Conoscenti, A. Vetro, and J. C. De Martin, “Blockchain for the
[31] Not Reported Not Reported 9 Internet of Things: A systematic literature review,” Proc. IEEE/ACS
(Van Der Poorten Int. Conf. Comput. Syst. Appl. AICCSA, no. April, 2017.
and Deschryver, [7] S. Huckle, R. Bhattacharya, M. White, and N. Beloff, “Internet of
Things, Blockchain and Shared Economy Applications,” Procedia
Comput. Sci., vol. 58, pp. 461–466, 2016.

174
[8] S. Madakam, R. Ramaswamy, and S. Tripathi, “Internet of Things Dehghan, and H. Rashidi, “A Framework for Cognitive Internet of
(IoT): A Literature Review,” J. Comput. Commun., vol. 03, no. 05, Things based on Blockchain,” in 4th International Conference on
pp. 164–173, 2015. Web Research, 2018.
[9] R. W. Thomas, D. H. Friend, L. A. Dasilva, A. B. Mackenzie, and V. [21] J. P. Koning, “Fedcoin : A Central Bank-issued Cryptocurrency,”
Tech, “Cognitive Networks : Adaptation and Learning to Achieve 2016. [Online]. Available:
End-to-End Performance Objectives,” no. December, pp. 51–57, https://static1.squarespace.com/static/55f73743e4b051cfcc0b02cf/t/5
2006. 8c7f80c2e69cf24220d335e/1489500174018/R3+Report-
[10] J. Patra, “Implementing blockchain for cognitive IoT applications , +Fedcoin.pdf. [Accessed: 13-Apr-2018].
Part 2 : Use vehicle sensor data to execute smart transactions in [22] T. Tata and B. Sweet, “Umbrella Coin : Peer-to-Peer Benefit
Blockchain Integrate the IBM Watson IoT Platform with IBM Packages,” 2017. [Online]. Available:
Blockchain services to develop a sample use case,” 2017. [Online]. https://icobazaar.com/static/0dba889e946117e54d29f1adc4b254c9/U
Available: https://www.ibm.com/developerworks/cloud/library/cl- MC.pdf. [Accessed: 13-Apr-2018].
blockchain-for-cognitive-iot-apps2/. [Accessed: 13-Apr-2018]. [23] Bitpark company, “CryptoCurrencies: Paving The Way to A New
[11] J. Manral, “IoT enabled Insurance Ecosystem - Possibilities Insurance Market CryptoCurrencies,” 2017. [Online]. Available:
Challenges and Risks,” pp. 1–18, 2015. https://bitpark.net. [Accessed: 13-Apr-2018].
[12] M. Raikwar, S. Mazumdar, S. Ruj, S. Sen Gupta, A. Chattopadhyay, [24] Etherisc, “Etherisc decentralized insurance,” 2018. [Online].
and K. Lam, “A Blockchain Framework for Insurance Processes,” Available: https://etherisc.com/. [Accessed: 13-Apr-2018].
2018. [25] IBM, “IBM Watson,” IBM, 2018. [Online]. Available:
[13] F. Lamberti, V. Gatteschi, C. Demartini, C. Pranteda, and V. https://www.ibm.com/watson/. [Accessed: 13-Apr-2018].
Santamaria, “Blockchain or not blockchain, that is the question of the [26] “Hyperledger Architecture , Volume II.” [Online]. Available:
insurance and other sectors,” IT Prof., 2017. https://www.hyperledger.org/wp-
[14] I. Nath, “Data Exchange Plattform to fight Insurance Fraud on content/uploads/2018/04/Hyperledger_Arch_WG_Paper_2_SmartCo
Blockchain,” IEEE 16th Int. Conf. Data Min. Work. Data, pp. 821– ntracts.pdf. [Accessed: 10-Jun-2018].
825, 2016. [27] “Hyperdedger,” The Linux Foundation, 2018. [Online]. Available:
[15] Y. Velner, J. Teutsch, and L. Luu, “Smart contracts make bitcoin https://www.hyperledger.org/.[Accessed: 12-June-2018].
mining pools vulnerable,” Lect. Notes Comput. Sci. (including [28] J. J. Bambara and P. R. Allen, Blockchain A Practical Guide to
Subser. Lect. Notes Artif. Intell. Lect. Notes Bioinformatics), vol. Developing Business, Law, and Technology Solutions. McGraw-Hill
10323 LNCS, pp. 298–316, 2017. Education, 2018.
[16] V. Gatteschi, F. Lamberti, C. Demartini, C. Pranteda, and V. [29] L. Chapman, -Henderson President, A. K. Ceo, and J. D. Rierson,
Santamaría, “Blockchain and Smart Contracts for Insurance: Is the “Understanding the Intersection of Resilience, Big Data, and the
Technology Mature Enough?,” Futur. Internet, vol. 10, no. 2, p. 20, Internet of Things in the Changing Insurance Marketplace,” 2015.
2018. [30] J. Forster and B. Entrup, “A Cognitive Computing Approach for
[17] E. Zanaj, K. Verushi, I. Enesi, and B. Zanaj, “Advances in Internet, Classification of Complaints in the Insurance Industry,” IOP Conf.
Data & Web Technologies,” vol. 17, pp. 394–402, 2018. Ser. Mater. Sci. Eng., vol. 261, no. 1, 2017.
[18] A. M. Saghiri and M. R. Meybodi, “An approach for designing [31] M. Van Der Poorten and P. Deschryver, “The Cognitive Insurer IBM
cognitive engines in cognitive peer-to-peer networks,” J. Netw. Industry Point View,” in Belgian Insurance Conference, 2017, pp. 1–
Comput. Appl., vol. 70, pp. 17–40, 2016. 23.
[19] K. Gholizadeh, A. M. Saghiri, M. Vahdati, M. Dehghan, and M. R. [32] P. L. Schwartz, “Analytics , Cognitive and IOT for Insurance Please
Meybodi, “A Framework for Cognitive Recommender Systems in the Note :,” 2016. [Online]. Available:
Internet of Things ( IoT ),” 2017 IEEE 4th Int. Conf. Knowledge- http://www.casact.org/education/infocus/2016/presentations/GS1_1.p
Based Eng. Innov., pp. 971–976, 2017. df. [Accessed: 13-Apr-2018].
[20] A. M. Saghiri, M. Vahdati, K. Gholizadeh, M. R. Meybodi, M.

175