Proceedings of the 52nd Hawaii International Conference on System Sciences | 2019

From Hype to Reality: A Taxonomy of Applications

Olga Labazova Tobias Dehling Ali Sunyaev University of Cologne Karlsruhe Institute of Technology Karlsruhe Institute of Technology [email protected] [email protected] [email protected]

Abstract Yet, challenges of developing blockchain-based systems outweigh envisioned benefits [8]. Most of Blockchain is a decentralized digital that the current blockchain projects could not move from challenges existing business models and theories by ideas to production use [13]. For example, projects shifting the trust from institutions towards aimed at employing to support algorithms. However, the number of successfully tokenization of space missions (e.g., SpaceBIT) or developed blockchain-based systems remains low. artificial intelligence [35] did not reveal proofs of This points towards a research gap between concept. Narrow-scoped blockchain prototypes blockchain applications and technical blockchain experience issues with scalability of blockchain characteristics. We answer the research question: protocols, waste of computational resources required What application areas fit blockchains with what for consensus mechanisms, traceability of users, and technical characteristics? We develop a taxonomy, a lack of network protection against fraud [11, 36, 44, which comprises six blockchain application areas 45]. Currently, practitioners continue experimenting that are classified across eight technical dimensions. with proofs of concept and system designs based on We demonstrate the utility of the taxonomy on ninety- trial-and-error approaches [13]. nine blockchain-based systems. We contribute to the Extant research in the blockchain domain is scientific literature by delimiting blockchain focused on the development of blockchain-based application areas, identifying new technical systems and the diversity of technical components dimensions, and linking application and technical (e.g., consensus mechanisms, permissioning) and knowledge on blockchain to guide development of applications (e.g., financial transactions, the internet blockchain-based systems. For practitioners, we of things). A closer examination of extant research present an overview of current blockchain-based reveals the diversity of blockchain application areas systems. with no-size-fits-all technical blockchain characteristics [20, 39, 44]. For example, the network is untrusted and requires a secure proof-of- work consensus mechanism [25] while a 1. Introduction business network ensures trust and can employ lighter consensus mechanisms, such as practical A blockchain is a decentralized digital ledger [12] Byzantine fault tolerance [19]. The relevant technical with the unique value proposition to shift the trust blockchain characteristics, however, remain abstract, from institutions towards algorithms [25]. The future fragmented, and scattered across applications. impact of blockchains on existing business models More knowledge connecting technical blockchain and theories might be comparable to the invention of characteristics and blockchain applications is crucial smart phones or the internet [3, 9, 27, 41]. Therefore, to provide the guidelines on development of researchers and practitioners jump on the blockchain successful blockchain-based systems. Trial-and-error bandwagon [1, 3] in attempts to replace established development leads to unfulfilled expectations in trust-based business models with blockchains [12, blockchain-based systems and loss of investments. 37]. The hype emerging around blockchains suggests Therefore, we answer the research question: What that blockchains can replace banks in the financial application areas fit blockchains with what technical sector [25, 37], support agreements among characteristics? individuals or internet-of-things devices using smart Taxonomies are used to organize knowledge in contracts [18, 28], and manage essential records (e.g., many fields (e.g., Darwin’s classification of species health records, education records) that are currently in biology) [4, 10, 24, 31]. We choose a taxonomy as maintained by centralized organizations [2, 32]. the fundamental tool to organize knowledge on blockchains [26]. We develop a taxonomy of

URI: https://hdl.handle.net/10125/59893 ISBN: 978-0-9981331-2-6 Page 4555 (CC BY-NC-ND 4.0)

blockchain applications, which captures six on reading and writing financial transactions to blockchain application areas that are classified across authorized users; decentralized payment platforms do eight technical dimensions [26]. The taxonomy is not require user authorization to read and write based on extant scientific literature, business reports, financial transactions. A review of and previous blockchain classifications. We investigates different consensus mechanisms, levels demonstrate the utility of the taxonomy by of anonymity, and data integrity among classifying ninety-nine blockchain-based systems cryptocurrencies [23]. Different consensus [12, 43]. Extant blockchain taxonomies and other mechanisms (e.g., proof-of-stake, practical Byzantine classifications describe blockchains from either fault tolerance) are determined to be suitable to technical or application perspectives [5, 17, 39, 44]. improve the efficiency of second-generation Our taxonomy is different because it integrates cryptocurrencies [6, 38, 45]. Compared to Bitcoin, technical and application knowledge that allows to Zerocoin guaranties a stronger anonymity of users guide the development of blockchain-based systems. that prevents user traceability [11, 14, 29] and This research contributes to the scientific has lower data integrity that allows for knowledge base in three ways. First, we establish an support of devices with low storage capacity (e.g., overview of extant research on blockchain mobile phones) [15]. Further overviews of key application areas. Second, we identify new technical technical characteristics of blockchains gather dimensions of importance to blockchain applications, previous findings in the financial sector including which complement extant work in the technical reading and writing permissions of transactions, literature. Third, we link blockchain application areas consensus mechanisms, anonymity levels, and other and technical blockchain characteristics, which can technical characteristics that are not focused on guide development of blockchain-based systems. For blockchain design but rather on interoperability (e.g., practitioners, the taxonomy gives an overview of chain modularity) [16, 17, 39, 44]. successful blockchain applications that can reduce Investigations of blockchain application areas development challenges for future blockchain-based start with the idea that blockchains can be useful systems. beyond the financial sector. Extant research focuses This manuscript proceeds as follows. We start predominantly on applying blockchains for digital with related research on blockchain. Next, we outline payments, certification, cloud storage, identity the approach employed for taxonomy development. management, energy distribution, and advanced Then, we present the taxonomy of blockchain tracking [30]. Business reviews of blockchain applications and demonstrate its utility on ninety-nine startups reveal new application areas including blockchain applications. Finally, we discuss principal customer loyalty, cybersecurity, digital rights findings, future research, limitations of our study, and management, digital voting and government, gaming, implications for theory and practice. content distribution, platform development, prediction markets, and smart contracts [12, 30]. 2. Related research Isolated knowledge of technical and application research causes hypes of blockchain application areas and technical blockchain characteristics. Further The scientific literature on blockchain is at an consideration and consolidation of application and early development stage. An absence of guidelines on technical knowledge on blockchains will result in a development of blockchain-based systems hinders foundational classification of blockchain application successful blockchain projects. Extant blockchain areas in alignment with technical blockchain taxonomies and other classifications consider characteristics and provide the first steps to guide the technical blockchain characteristics and blockchain development of successful blockchain-based systems. application areas separately. Technical blockchain classifications are focused on the diversity of technical components (e.g., permissions to read 3. Research approach transactions, consensus mechanisms) and cover predominantly the financial sector [6, 21, 23, 38, 43, To organize knowledge on blockchains, we use 45]. For instance, a study comparing digital payment the method for taxonomy development proposed by providers identifies permissions to read and write Nickerson et al., who define a taxonomy as a set of financial transactions as important technical dimensions [26]. Each dimension consists of characteristics to consider when choosing between “mutually exclusive and collectively exhaustive centralized and decentralized payment platforms characteristics in a way that each object under [21]. Centralized payment platforms give permissions consideration has one and only one” [26:5]

Page 4556

characteristic in every dimension. The taxonomy ” on October 17, 2016, in title, development method proceeds in three stages (Figure abstract, and keywords, covering the whole period of 1). In the initial stage, a metacharacteristic and publications [7, 40]. The search returned fifty-one ending conditions are defined according to the papers. After screening of titles and abstracts, we purposes of the taxonomy to be developed. In the coded the forty-one remaining relevant articles. In the main stage, the taxonomy is developed. Taxonomy first iteration, we identified six dimensions with objects (here application cases), dimensions, and fourteen characteristics and six application areas with characteristics are identified during inductive or sixteen application cases. The analysis of the deductive iterations. In inductive iterations, empirical scientific literature revealed detailed information on cases are analyzed to determine dimensions and separate blockchain characteristics (e.g., consensus characteristics in the taxonomy. In deductive mechanisms) or specific blockchain application iterations, dimensions and characteristics are derived examples (e.g., energy markets, prediction platforms) from the scientific knowledge base. In the final stage, but lacked comprehensiveness. In the second the taxonomy is evaluated against ending conditions. iteration, we analyzed business reviews, which provide less profound but more comprehensive Deductive and inductive information. We investigated twenty business reports iterations by national agencies, consulting companies, and

;

international institutions. We revised the taxonomy

:

: and added two dimensions, seven characteristics, and Are Yes nine application cases. The third iteration was

iteration 2 iteration ending classifications conditions deductive, where we derived characteristics, met? dimensions, and application cases from fifteen

Nickerson et al.

A taxonomy of

Inductive iteration Inductive 1: iteration Inductive

Metacharacteristic

Deductive iterationDeductive 3

reviewing reviewing

reviewing business reports reviewing business

reviewing reviewing scientific papers previous classifications. We used all previous

information processing

all all ending conditions by No

blockchain applications classifications that we could identify in extant Figure 1. Research approach for development of the literature until May 2018. Our taxonomy covers all taxonomy of blockchain applications. characteristics in classifications related to technical blockchain characteristics. All ending conditions proposed by Nickerson at 3.1. Development of the taxonomy of al. [26] were fulfilled after the third iteration as blockchain applications follows. First, all found blockchain application cases described in the scientific literature or business The objective of the taxonomy is to classify reports can be classified into an application case in blockchain application areas based on technical the taxonomy. Second, each dimension is unique and blockchain characteristics. Therefore, we selected mutually exclusive, and each characteristic is unique technical blockchain characteristics (e.g., consensus within its dimension. Third, all application cases mechanism, anonymity level) as the were classified with a single characteristic for each metacharacteristic. The choice and combination of dimension. Fourth, the taxonomy is concise— technical blockchain characteristics are central to the consists only of meaningful dimensions that classify success or failure of blockchain-based systems. The application cases. Fifth, the taxonomy is robust— metacharacteristic serves as basis for identification of differentiates each application case from all others. further dimensions and characteristics. Sixth, the taxonomy is explanatory, comprehensive, We developed the taxonomy in three iterations. and extensible—highlights the main features of each The first two iterations were inductive iterations, application case and can be extended when new where we identified application cases to derive application cases arise. dimensions and characteristics. For each inductive iteration, we used different types of sources: 3.2. Data analysis scientific literature and business reviews, respectively. The third iteration was a deductive iteration where we revised the taxonomy based on To analyze the sources, we used three types of previous classifications. In the first iteration, we coding: open coding, axial coding, and selective searched articles in the web of science core collection1 with the search string “blockchain OR Citation Index (1975-present), Conference Proceedings Citation Index- Science (1990-present), Conference Proceedings Citation Index- Social Science & Humanities (1990-present), Book Citation Index– Science (2005-present), Book Citation Index– Social 1 Used indices: “Science Citation Index Expanded (1900-present), Sciences & Humanities (2005-present), and Emerging Sources Social Sciences Citation Index (1900-present), Arts & Humanities Citation Index (2015-present)”

Page 4557

coding [33, 42]. Open coding is a process for application cases and usually include several highly grouping categories and subcategories [33:12]. Axial trusted nodes for arriving at system-level agreements. coding is a process for testing “that categories are The fourth dimension is anonymity level and assesses related to their subcategories, and the relationships whether users can be matched to identities. If against data” [33:13]. Selective coding is a process blockchains have the characteristic anonymous, users “by which all categories are unified around a ‘core’ do not have to provide any data to work with category, and categories that need further explication blockchains. If blockchains are pseudonymous, users are filled-in with descriptive details” [33:14]. We have to work under a pseudonym. Blockchains with applied open coding for initial categorization of the characteristic identifiable ask for or automatically dimensions, characteristics, application areas, and collect personally identifiable information, such as application cases; axial coding for removal of email addresses. The fifth dimension is event overlapping dimensions, characteristics, application handling and discerns whether blockchains can areas, and application cases while iteratively testing handle application logic or events. No event handling the taxonomy against data; and selective coding to shows an inability to handle application logic. Fixed classify each application case with a characteristic for event handling supports built-in events. Custom event each dimension. One researcher coded the sources handling means that a blockchain supports processing three times, in November 2016, April 2017, and of any application logic provided by users. The sixth November 2017, and other researchers validated the dimension is data exchange type that focuses on the results after each iteration [34]. Disputes were type of information sharing between users on resolved in group discussions. blockchains and includes the characteristics transaction and content. Transaction implies an 4. Taxonomy of blockchain applications exchange of logs of executed actions. Content means that digital assets, such as documents, messages, and video or music files, are exchanged. The seventh The developed taxonomy consists of eight dimension is encryption and specifies whether data dimensions with twenty-one technical characteristics on blockchains is encrypted. Unencrypted means that and six application areas with twenty-five application no data on the blockchain is encrypted. Partially- cases (Table 1). encrypted represents blockchain, where some data is encrypted. Totally-encrypted means that all data on 4.1. Technical blockchain characteristics blockchains is encrypted and has to be decrypted for all operations. The eighth dimension is history The first dimension is reading access and retention and ascertains whether the whole represents different modes for reading information on blockchain or only its recent updates are kept and blockchains. Private reading allows only authorized distributed between hosts. Whole retention means members to access a blockchain. Public reading that the whole history starting with a genesis block is access allows everyone to read data from a kept in a blockchain and distributed between nodes. blockchain. The second dimension is writing access Recent updates retention specifies that only the latest and represents different modes of writing information updates are kept and distributed. on a blockchain. Permissioned writing access requires users to be authorized to add transactions. If 4.2. Blockchain application cases writing access is unpermissioned, a user does not have to be authorized to add transactions. The third We identified six blockchain application areas dimension is main consensus mechanism and is comprising a total of twenty-five application cases. concerned with employed means for updating Application areas capture the basic functionalities blockchains; we focus on four predominant that can be performed by blockchains and group consensus mechanisms. Proof-of-work requires some application cases with similar semantic features and resources (or work) from a requester, usually similar combinations of technical blockchain processing time of a computer to solve a characteristics. The first application area is financial computationally difficult puzzle. Proof-of-stake asks transactions and captures seven application cases users to proof the ownership of a certain amount of concerned with money transfer and exchange. digital data to establish their stake in this data. Conventional cryptocurrencies use public Practical Byzantine fault tolerance gathers individual unpermissioned blockchains, where consensus is decisions made by trusted nodes in a network that achieved through proof-of-work, and users act under together determine system-level agreements. Self- pseudonyms. Blockchains with the same developed consensus mechanisms are used in some characteristics except for anonymous user access

Page 4558

support anonymous cryptocurrencies. To confirm enterprise asset management applies private interest of users in blockchain and to reduce permissioned blockchains that reach system-level processing costs, wealth storage & micro-payments consensus by practical Byzantine fault tolerance and require proof-of-stake consensus mechanisms along require unique identification of nodes. with public unpermissioned blockchains and The fourth application area is storage and is pseudonymous users. Public permissioned concerned with keeping digital assets, such as blockchains with some modifications of proof-of- certificates or music and video files, on blockchains. work consensus mechanism support financial Open access publishing uses public blockchains and services by expanding the functionality of payments requires no data encryption. Content preview through financial checks and deposits. Energy- employs public blockchains with partial encryption efficient financial services use blockchains with the of data. Blockchain-based decentralized storage is same characteristics as financial services except for implemented on public blockchains with total data proof-of-stake consensus mechanisms. Enterprise encryption and some modifications for faster content global- and micro financial transactions employ sharing and decoding. private unpermissioned blockchains with practical The fifth application area is communication. Byzantine fault tolerance consensus mechanism, Broadcasting is supported by public unpermissioned which requires unique identification of nodes in the blockchains with proof-of-work consensus network. Global centrally issued financial mechanisms and without data encryption because the instruments are deployed on private permissioned content is intended for mass communication. Public blockchains with self-developed consensus permissioned blockchains with proof-of-work mechanisms, which also require unique identification consensus mechanisms are suitable for discussion of the nodes. forums, which allow any user to participate in The second application area is smart contracts communication but automatically collect IP and processes application logic. The application area addresses. Internet-of-things communication uses contains eight application cases. Most smart contracts private unpermissioned blockchains and practical work on public unpermissioned blockchains with a Byzantine fault tolerance consensus mechanism to proof-of-work consensus mechanism. At the same control information exchange between devices in time, a proof-of-stake consensus mechanism supports enterprise or home networks. energy-efficient smart contracts. For testing purposes, The sixth application area is ranking with a single one can create private blockchains that comprise only application case. Global reputation & rating is one node. Community smart contracts, which must supported by public permissioned blockchain with comply with different community rules, are based on proof-of-work consensus mechanisms and automatic public permissioned blockchains with proof-of-work collection of identifiers to link identities to individual consensus mechanisms. Energy-efficient community users and to prevent users from obtaining more than smart contracts apply proof-of-stake consensus one identity. mechanisms. Enterprise smart contracts use private unpermissioned blockchains. Global agreements 4.3. Demonstration of the utility of the between institutions can be achieved based on private taxonomy permissioned blockchains. The third application area is data management We demonstrate the utility of the taxonomy on and is concerned with information management, such ninety-nine blockchain-based systems mentioned in as authentication, know-your-customer services, and the scientific and business sources. To classify control of business assets. The area includes three identified blockchain-based systems with the application cases. To manage assets registered off- taxonomy, we used white papers, the systems’ chain, global authentication and ownership require websites, press releases, and set up the systems and public unpermissioned blockchains with proof-of- tested them if it was possible. The demonstration of work consensus mechanisms and pseudonymous the utility of the taxonomy shows that the taxonomy users. Sharing economies and enterprise asset classifies successful blockchain-based systems and management require data management with purposefully does not classify some blockchain-based identification and authorization schemes systems. implemented directly on a blockchain. To avoid fraud although opening a network for many nodes, sharing 4.3.1. Classified blockchain-based systems. The economies use public permissioned blockchains with gathered blockchain-based systems predominantly proof-of-work consensus mechanisms and cover the financial sector. Anonymous identifiable users. To keep information confidential, cryptocurrencies include Zerocoin, Darkcoin,

Page 4559

CryptoNote, and . Conventional support enterprise global and micro- financial cryptocurrencies comprise Bitcoin, , transactions. R3, Fedcoin, Symbiont Assembly, Litecoin, , DagCoin, Crypt Cryptosigma, RSCoin, and Onecoin represent global centrally DigixGlobal, GameCredits, Bitpay, and SolarCoin. issued financial instruments. , Navcoin, AML, and Blackcoin target Smart contracts are popular for the identified wealth storage & micro-payments. , blockchain-based systems. , Hawk, Stratis, Mastercoin, and DigitalNote execute financials Qtum, Blockcypher, Deckbound, Rootstock, iExec, services. BitShares allows energy-efficient financial Chimera, WeTrust, Sia, and Maidsafe support services. Ripple, SWIFT gpi, Stellar, and BitPesa original smart contracts. Testing of smart contracts is

Table 1. A taxonomy of blockchain applications.

Data A TAXONOMY OF BLOCKCHAIN Reading Writing Main consensus Anonymity Event exchange Encryption History APPLICATIONS access access mechanism level handling type retention Pr Pu P U W S B SD A P I N F C T C U P T W R Anonymous cryptocurrencies X X X X X X X X Cryptocurrencies X X X X X X X X Wealth storage & micro-payments X X X X X X X X Financial services X X X X X X X X Financial Energy-efficient X X X X X X X X transactions financial services Enterprise global and micro- financial X X X X X X X X transactions Global centrally issued financial X X X X X X X X instruments Smart contracts X X X X X X X X Testing of smart contracts X X X X X X X X Energy-efficient smart contracts X X X X X X X X Testing of energy-efficient X X X X X X X X smart contracts Community Smart contracts smart contracts X X X X X X X X Energy-efficient community smart X X X X X X X X contracts Enterprise smart contracts X X X X X X X X Global agreements between X X X X X X X X institutions Global authentication X X X X X X X X and ownership Data management Sharing economies X X X X X X X X Enterprise asset management X X X X X X X X Open access publishing X X X X X X X X Storage Content preview X X X X X X X X Decentralized storage X X X X X X X X Broadcasting X X X X X X X X Discussion Communication Forum X X X X X X X X IoT communication X X X X X X X X Global reputation Ranking & rating X X X X X X X X

LEGEND Event handling No – No: blockchain does not support any events X – characteristics belong to an application case F – Fixed: blockchain supports built-in events C – Custom: blockchain supports processing of events created by user Reading access Data exchange type Pr – Private: only authorized members of a limited community can read blockchain T – Transaction: logs of actions executed are exchanged among users and Pu – Public: everybody can read a blockchain recorded on a blockchain Writing access C – Content: digital assets are exchanged among users and recorded on a P – Permissioned: a user should be authorized to validate transactions blockchain U – Unpermissioned: a user can validate transactions without authorization Encryption Main consensus mechanism U – Unencrypted: all data on a blockchain is unencrypted W – Proof-of-work: consensus for secure blockchain updating is achieved by Proof-of-Work P – Partially-encrypted: data on a blockchain is partially encrypted S – Proof-of-stake: consensus for secure blockchain updating is achieved by Proof-of-Stake T – Totally-encrypted: all data on a blockchain is encrypted B – Practical Byzantine fault tolerance: secure blockchain updating is achieved by History retention agreements of trusted nodes W – Whole: blockchain keeps whole transaction history from a genesis SD – Self-developed mechanism: consensus for secure blockchain updating is achieved by block self-developed mechanism R – Recent updates: blockchain keeps only recent updates of the transaction Anonymity level history A – Anonymous: users do not have to provide any data for working with blockchain P – Pseudonymous: users can work with a blockchain under a pseudonym I – Identifiable: users should provide personal data to work with a blockchain

Page 4560

possible on Ethereum (testing environment), Hawk can be replaced by conventional peer-to-peer (testing environment), and EOS. Casper, Tendermint, systems). and develop energy-efficient smart contracts. The second reason of unclassified blockchain- Testing of energy-efficient smart contracts is based systems is combinations of technical performed on Casper (testing environment). blockchain characteristics that appear to be Counterparty supports community smart contracts. ineffective. These blockchain-based systems exhibit Lisk and execute energy-efficient community or intensify security threats or privacy concerns. For smart contracts. Hyperledger, Ripple Codius, Eris example, hackers attack blockchains by forking them, (Monax), Digital asset, Waves, and Catenis developers of blockchain-based systems can falsify Enterprise support enterprise smart contracts. R3 data on blockchains, and users can be traceable when Codra allows reaching global agreements between permissions to read and write data on blockchains do institutions. not comply with consensus mechanisms or with Data management on blockchains gains anonymity protection of users. momentum. Colored coins, , Onename, The third reason of unclassified blockchain-based POEX.IO, OP_RETURN, Everpass, The Real systems is a combination of blockchain application McCoy, BitHealth, BitAuth, UniquID, NEM areas and technical blockchain characteristics that Apostille, Blockname, Filament, ePlug, and Shocard appear to be unsuitable. For example, a blockchain- represent global authentication and ownership. based system that aims to manage certificates Iconomi, NEO, Ridde & code, Aragon, and La’Zooz between trustful organizations (e.g., school diplomas are examples of sharing economies. Everledger, between schools and employee companies) is an PeerNova, Factom, Chromaway, BlockVerify, example of enterprise asset management. However, PeerNova, Chronicled, and ShoBadge support an application we identified uses a public blockchain enterprise asset management. with a proof-of-work consensus mechanism instead A smaller number of blockchain applications of a private blockchain with a practical Byzantine supports blockchain-based storage. Synereo fulfills fault tolerance consensus mechanism. The reason open access publishing. Kishigami et al. [22] why the application uses a blockchain is not due to describe content preview on blockchains; although the actual number of nodes but due to the borrowed we did not find blockchain-based systems to support public infrastructure. The following concerns arise. If the application case, we decided to keep the the application uses a public blockchain, transactions application case for further research. The Storj are expensive because of the consensus mechanism. project examines decentralized storage on For transactions on this blockchain the issuers of the blockchain. certificates (e.g., schools) must be trustful to prevent Communication is not often implemented on information manipulation or fraud (e.g., an actor blockchains. Basic Attention Token shows could send transactions to himself to change records). broadcasting. Blockchain-based discussion forums However, if issuers are trustful, a public blockchain include Whisper and Matchpool. Blockchain of is useless. Therefore, the blockchain application Things and IBM Adept support internet-of-things ignores the main dilemma in using blockchains and communication. public-private infrastructure: the more trustful issuers Ranking on blockchains is an uncommon are, the less energy-consuming the employed blockchain application case. Augur, TRST.im, The consensus mechanism should be. World Table, and TrustDavis support global reputation & rating. 5. Discussion

4.3.2. Unclassified blockchain-based systems. We The developed taxonomy serves as a bridge found blockchain-based systems that purposefully between blockchain technology and blockchain remain unclassified by our taxonomy. The first application areas. The taxonomy constitutes a tool to reason for unclassified blockchain-based systems is connect technical blockchain characteristics across a an application area that appears to be unsuitable for range of foundational application cases. There are blockchains. Such blockchain applications have five principal findings. First, application areas in the broad ideas and aim to replace current information taxonomy are at different maturity levels. Financial systems with blockchains (e.g., decentralized transactions constitutes the most mature application internet); however, they do not result in any proofs of area and is supported by existing proofs of concept. concept. Other examples arise when blockchain Smart contracts have found much attention because applications use blockchains when blockchains are of the idea to execute agreements on blockchains not needed (e.g., private messengers on blockchains instead of third parties. Data management gains

Page 4561

momentum because of emerging application cases technical characteristics that contradict the taxonomy (e.g., enterprise asset management on private can lead to inefficient technical designs. blockchains). Storage, communication, and ranking Inconsistencies between application areas and on blockchains are less prevalent. Blockchain technical designs may indicate a lack of compliance scalability issues prevent storage of data on with technical and application requirements. blockchains. The value of applying blockchains for However, the taxonomy is only based on extant communication and ranking is specific to each knowledge in research and practice and this assertion application case. In particular, it is challenging to requires further research. support mobile devices when energy-consuming There are three promising areas for future consensus mechanisms and the transfer of the whole research. First, research that replicates our research transaction history are required. approach with more or different scientific and Second, application cases inside one application business sources will be useful to falsify or area vary in the dimensions reading access, writing corroborate our findings. Second, further analysis of access, main consensus mechanism, and anonymity theoretical findings allows to hypothesize about the level. The characteristics in these dimensions depend relationships between application areas and technical on the required levels of for blockchain characteristics. Third, research that application cases. The more centralization is focuses on socio-economic concepts different from required, the more private reading access and the application areas, for example, market regulations in more permissioned writing access is required. Main different countries will be useful to contextualize the consensus mechanism and anonymity level follow taxonomy for different industries and domains. the required level of decentralization so that the more This study is not without limitations. First, the centralization is required, the less energy-consuming taxonomy cannot identify application areas that may are consensus mechanisms and the less anonymous emerge in the future. The rapidly evolving nature of are nodes. the blockchain domain will necessitate an extension Third, to classify application areas, we reveal new of the taxonomy with new application cases. Second, technical dimensions that are overlooked in extant the identified application areas do not directly capture technical classifications on blockchains due to its more complex services, such as prediction markets or predominant focus on the financial sector. The new crowdsourcing platforms; instead, we decided to dimensions are event handling, data exchange type, break complex application cases down into the basic encryption, and history retention. Custom event functionalities that can be performed by blockchains. handling specifies smart contracts. Data exchange This research contributes to the scientific type allocates whether data is stored on or off literature on blockchain in three ways. First, blockchains. Encryption is different between allocation of blockchain application cases based on application cases that require to store content or technical blockchain characteristics reduces the hype transactions on blockchains. History retention is around blockchain application possibilities. A different for application cases that store blockchains classification of application areas that, along with on small-capacity external devices and have to semantic differences, is based on technical remove old information from blockchains. characteristics make the identification of application Fourth, not all and different technical blockchain areas more meaningful. The well-studied financial characteristics are suitable for different application sector can serve as a good example for how to areas. For example, communication systems based on leverage blockchains in less studied application areas private permissioned blockchains do not appear to and the other application areas may reveal create additional value compared to peer-to-peer opportunities that have been overlooked in the messengers such as Telehash, which are used by financial sector. Second, we identified additional many decentralized services (e.g., IBM Adept). technical dimensions of importance in the blockchain However, this statement requires further domain. While some of the taxonomy dimensions investigation. (reading access, writing accesses, main consensus Fifth, the taxonomy purposefully avoids the mechanisms, and anonymity level) align with classification of poorly developed blockchain-based previous taxonomies, the remaining dimensions systems because blockchain application cases are (event handling, data exchange type, encryption, and identified and related to unique and effective history retention) represent specific application areas combinations of technical characteristics. Therefore, and complement previous taxonomies by offering blockchain-based systems that are not captured by the more comprehensive insights into the technical taxonomy might represent application areas that are nature of blockchains. Therefore, technical research unsuitable for blockchains. Combinations of can go beyond the Bitcoin blockchain and focus on

Page 4562

other areas, for example, development of a taxonomy accounts for twenty-five application cases blockchain-based protocol for data transmission in aggregated into six application areas that relate to healthcare. Third, previous taxonomies consider twenty-one technical blockchain characteristics in technical knowledge [17] or application knowledge eight dimensions. Overall, the taxonomy consolidates [12] separately. Our taxonomy combines the extant knowledge on blockchains to calm the knowledge, which allows to bridge the gap between blockchain hype and foster development of more extant technical and application research streams on realistic blockchain-based systems. blockchain. Linking application areas and technical characteristics informs step-by-step guidelines for 7. References leveraging blockchains across application areas. Such guidelines might be useful for further development of [1] Avital, M., J.L. King, R. Beck, M. Rossi, and R. successful blockchain-based systems. Teigland, “Jumping on the Blockchain Bandwagon : This research contributes to practice in three Lessons of the Past and Outlook to the Future Panel”, ways. First, we present further evidence that Thirty Seventh International Conference on Information blockchains are not only applicable to the financial Systems, (2016), 1–6. sector, which is the focus of the majority of [2] Azaria, A., A. Ekblaw, T. Vieira, and A. Lippman, blockchain projects but also for other promising “MedRec: Using Blockchain for Medical Data Access and areas. Thus, other industries can use blockchain Permission Management”, 2nd International Conference on advantages for resolving their challenges. For Open and Big Data (OBD), (2016), 25–30. example, in the media industry blockchain-based data [3] Beck, R., and C. Müller-Bloch, “Blockchain as Radical Innovation: A Framework for Engaging with Distributed management may be useful to monitor the use of Ledgers as Incumbent Organization”, 50th Hawaii media content to prevent copyright infringements. International Conference on System Sciences, (2017), Second, we highlight other blockchain characteristics 5390–5399. besides the widely-known public blockchains that [4] Bloom, B., “Bloom ’ s Taxonomy”, ReVision, 2001, pp. can be useful if public blockchains cannot be 1–6. employed. Businesses may consider implementation [5] Böhm, M., J. Weking, F. Fortunat, S. Müller, and I. of private blockchains that store information in a Welpe, “The Business Model DNA : Towards an Approach more reliable way. Third, we have proposed the for Predicting Business Model Success”, 13th International taxonomy of blockchain applications to guide Conference on Wirtschaftsinformatik, (2017), 9–11. [6] Bonneau, J., A. Miller, J. Clark, A. Narayanan, J.A. development of more successful blockchain-based Kroll, and E.W. Felten, “Research Perspectives and systems. The taxonomy establishes an overview of Challenges for Bitcoin and Cryptocurrencies”, IEEE blockchain applications, organizes them in Symposium on Security and Privacy, 2015, pp. 104–121. application areas, and relates them to technical [7] vom Brocke, J., A. Simons, B. Niehaves, et al., blockchain characteristics. Furthermore, the “Reconstructing the Giant: On the Importance of Rigour in taxonomy can be used to avoid poorly designed Documenting the Literature Search Process”, 17th blockchain applications. This might be useful for European Conference on Information Systems 9, 2009, pp. practitioners to identify the more promising 2206–2217. blockchain projects and assess risks during [8] Condos, J., W.H. Sorrell, and S.L. Donegan, “Blockchain Technology: Opportunities and Risks”, blockchain implementation. For example, chief Vermont(Vermont), 2016, pp. 1–35. information officers could learn which modules in the [9] Constantinides, P., O. Henfridsson, and G. Parker, enterprise information systems landscape can be “Platforms and Infrastructures in the Digital Age”, realized on blockchains and developers could learn Information Systems Research(May), 2018, pp. 1–20. which peer-to-peer system prototypes are worth to be [10] Darwin, C., “The Origin of Species (1859)”, In develop on blockchains. Foundations of Psychological Thought: A History of Psychology. 2009, 320–336. [11] Fabian, B., T. Ermakova, and U. Sander, “Anonymity 6. Conclusion in Bitcoin – The Users’ Perspective”, 37th International Conference on Information Systems (ICIS 2016), (2016), 1– A blockchain is a decentralized digital ledger with 12. a largely untapped potential to enhance many aspects [12] Friedlmaier, M., A. Tumasjan, and I.M. Welpe, in the information systems domain. Currently, “Disrupting Industries With Blockchain: The Industry, research streams on blockchain remain disconnected, Venture Capital Funding, and Regional Distribution of which prevents further development of successful Blockchain Ventures”, SSRN Electronic Journal, 2016. [13] Furlonger, D., and R. Valdes, Practical Blockchain: A blockchain-based systems. Our work consolidates Gartner Trend Insight Report, 2017. knowledge on technical blockchain characteristics [14] Garman, C., M. Green, I. Miers, and A.D. Rubin, and application areas in the form of a taxonomy. The

Page 4563

“Rational Zero: Economic Security for Zerocoin with Privacy, Security, Risk, and Trust, and IEEE International Everlasting Anonymity”, International Conference on Conference on Social Computing, (2011), 1318–1326. Financial Cryptography and Data Security, Springer, [30] Salviotti, G., L.M. De Rossi, and N. Abbatemarco, “A Berlin, Heidelberg, 140–155. Structured Framework to Assess the Business Application [15] Gibbs, T., and S. Yordchim, “Thai Perception on Landscape of Blockchain Technologies”, Proceedings of Litecoin Value”, International Journal of Social, the 51st Hawaii International Conference on System Behavioral, Educational, Economic, Business and Sciences, (2018), 3467–3476. Industrial Engineering 8(8), 2014, pp. 2613–2615. [31] Schneider, S., J. Lansing, F. Gao, and A. Sunyaev, “A [16] Glaser, F., “Pervasive Decentralisation of Digital Taxonomic Perspective on Certification Schemes: Infrastructures: A Framework for Blockchain enabled Development of a Taxonomy for Cloud Service System and Use Case Analysis”, 50th Hawaii International Certification Criteria”, Proceedings of the Annual Hawaii Conference on System Sciences, (2017), 1543–1552. International Conference on System Sciences, (2014), [17] Glaser, F., and L. Bezzenberger, “Beyond 4998–5007. Cryptocurrencies - A Taxonomy of Decentralized [32] Sharples, M., and J. Domingue, “The Blockchain and Consensus Systems”, Proceedings of the 23rd European Kudos: A Distributed System for Educational Record, Conference on Information Systems (ECIS 2015), (2015), Reputation and Reward”, Lecture Notes in Computer 1–18. Science (including subseries Lecture Notes in Artificial [18] Higgins, S., “Factom Partners With Honduras Intelligence and Lecture Notes in Bioinformatics), (2016). Government on Blockchain Tech Trial”, CoinDesk, 2015. [33] Strauss, A., and J. Corbin, Strauss, A., & Corbin, J. https://www.coindesk.com/factom-land-registry-deal- (1990)., 2008. honduran-government/ [34] Strauss, A.L., Qualitative Analysis for Social [19] Hyperledger Architecture Working Group, Scientists, 1987. “Hyperledger Architecture , Volume 1”, Hyperledger.org [35] Swan, M., “Blockchain Thinking : the Brain as a 1, 2017, pp. 15. Decentralized Autonomous Corporation [Commentary]”, [20] Kannengießer, N., S. Lins, T. Dehling, and A. IEEE Technology and Society Magazine 34(4), 2015, pp. Sunyaev, “What Does Not Fit Can be Made to Fit ! Trade- 41–52. Offs in Distributed Ledger Technology Designs”, [36] Swan, M., Blockchain: Blueprint for a New Economy, Proceedings of the 52nd Hawaii International Conference 2015. on System Sciences, (2019). [37] The Economist, “The Trust Machine - The promise of [21] Kazan, E., C. Tan, and E.T.K. Lim, “Towards a the blockchain”, The Economist, 2015, 2–4. Framework of Digital Platform Disruption : A Comparative [38] Tschorsch, F., and B. Scheuermann, “Bitcoin and Study of Centralized & Decentralized Digital Payment Beyond: A Technical Survey on Decentralized Digital Providers”, 25th Australasian Conference on Information Currencies”, IEEE Communications Surveys and Tutorials, Systems(2010), 2014, pp. 10. (2016). [22] Kishigami, J., S. Fujimura, H. Watanabe, A. [39] Walsh, C., P. O’Reilly, R. Gleasure, J. Feller, L. Nakadaira, and A. Akutsu, “The Blockchain-Based Digital Shanping, and J. Cristoforo, “New Kid on the Block: a Content Distribution System”, 2015 IEEE Fifth Strategic Archetypes Approach to Understanding the International Conference on Big Data and Cloud Blockchain”, 37th International Conference on Information Computing, (2015), 187–190. Systems (ICIS), (2016), 1–12. [23] Morisse, M., “Cryptocurrencies and Bitcoin: Charting [40] Webster, J., and R.T. Watson, “Analyzing the Past to the Research Landscape”, Americas Conference on Prepare for the Future: Writing a Literature Review.”, MIS Information Systems, 2015. Quarterly 26(2), 2002, pp. xiii–xxiii. [24] Mrosek, R., T. Dehling, and A. Sunyaev, “Taxonomy [41] Welpe, I., L. Zavolokina, H. Krcmar, and M. Müller, of Health IT and Medication Adherence”, Health Policy “Introduction to Distributed Ledger Technology , and Technology 4(3), 2015, pp. 215–224. Blockchain Minitrack”, (2018), 3465–3466. [25] Nakamoto, S., “Bitcoin: A Peer-to-Peer Electronic [42] Wiesche, M., M.C. Jurisch, P.W. Yetton, and H. Cash System”, Www.Bitcoin.Org, 2008, pp. 9. Krcmar, “Grounded Theory Methodology in Information [26] Nickerson, R.C., U. Varshney, and J. Muntermann, “A Systems Research”, MIS Quarterly 41(3), 2017, pp. 685- Method for Taxonomy Development and its Application in A9. Information Systems”, European Journal of Information [43] Wörner, D., T. Von Bomhard, Y.-P. Schreier, and D. Systems 22(3), 2013, pp. 336–359. Bilgeri, “The Bitcoin Ecosystem: Disruption beyond [27] Pongnumkul, S., C. Siripanpornchana, and S. financial Services?”, Twenty-Fourth European Conference Thajchayapong, “Performance Analysis of Private on Information Systems (ECIS), (2016), 1–16. Blockchain Platforms in Varying Workloads”, 2017 26th [44] Xu, X., I. Weber, M. Staples, et al., “A Taxonomy of International Conference on Computer Communications Blockchain-Based Systems for Architecture Design”, and Networks, ICCCN 2017, (2017). Proceedings - 2017 IEEE International Conference on [28] Pureswaran, V., S. Panikkar, S. Nair, and P. Brody, Software Architecture, ICSA 2017, (2017), 243–252. Empowering the Edge: Practical Insights on a [45] Yli-Huumo, J., D. Ko, S. Choi, S. Park, and K. Decentralized Internet of Things, 2015. Smolander, “Where is Current Research on Blockchain [29] Reid, F., and M. Harrigan, “An Analysis of Anonimity Technology? - A Systematic Review”, PLoS ONE 11(10), in the Bitcoin System”, IEEE International Conference on 2016, pp. 1–27.

Page 4564