Academic Paper on Steemit & DPOS from Cornell University! arXiv.org > cs > arXiv:1904.07310 [Submitted on 15 Apr 2019] Incentivized Blockchain-based Social Media Platforms: A Case Study of Steemit Chao Li, Balaji Palanisamy
OK WOW I just Found a CORNELL UNIVERSITY @redes ? ) Study on Steemit from Apr 2019!
https://arxiv.org/pdf/1904.07310.pdf
arXiv.org > cs > arXiv:1904.07310
Search...
Help | Advanced Search
All fields
Search
Computer Science > Social and Information Networks
[Submitted on 15 Apr 2019]
Incentivized Blockchain-based Social Media Platforms: A Case Study of Steemit
Chao Li, Balaji Palanisamy
https://arxiv.org/abs/1904.07310
This paper presents an empirical analysis of Steemit, a key representative of the emerging incentivized social media platforms over Blockchains, to understand and evaluate the actual level of decentralization and the practical effects of cryptocurrency-driven reward system in these modern social media platforms. Similar to Bitcoin, Steemit is operated by a decentralized community, where 21 members are periodically elected to cooperatively operate the platform through the Delegated Proof-of-Stake (DPoS) consensus protocol. Our study performed on 539 million operations performed by 1.12 million Steemit users during the period 2016/03 to 2018/08 reveals that the actual level of decentralization in Steemit is far lower than the ideal level, indicating that the DPoS consensus protocol may not be a desirable approach for establishing a highly decentralized social media platform. In Steemit, users create contents as posts which get curated based on votes from other users. The platform periodically issues cryptocurrency as rewards to creators and curators of popular posts. Although such a reward system is originally driven by the desire to incentivize users to contribute to high-quality contents, our analysis of the underlying cryptocurrency transfer network on the blockchain reveals that more than 16% transfers of cryptocurrency in Steemit are sent to curators suspected to be bots and also finds the existence of an underlying supply network for the bots, both suggesting a significant misuse of the current reward system in Steemit. Our study is designed to provide insights on the current state of this emerging blockchain-based social media platform including the effectiveness of its design and the operation of the consensus protocols and the reward system.
Comments: to appear in ACM WebSci 2019
Subjects: Social and Information Networks (cs.SI); Cryptography and Security (cs.CR)
Cite as: arXiv:1904.07310 [cs.SI]
(or arXiv:1904.07310v1 [cs.SI] for this version)
Bibliographic data
[Enable Bibex (What is Bibex?)]
Submission history
From: Chao Li [view email]
[v1] Mon, 15 Apr 2019 20:05:40 UTC (4,503 KB)
Download:
PDF
Other formats
(license)
Current browse context:
cs.SI
< prev | next >
new | recent | 1904
Change to browse by:
cs
cs.CR
References & Citations
NASA ADS
Google Scholar
Semantic Scholar
DBLP - CS Bibliography
listing | bibtex
Chao Li
Balaji Palanisamy
Export citation
Bookmark
BibSonomy logo Mendeley logo Reddit logo ScienceWISE logo
Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)
https://arxiv.org/pdf/1904.07310.pdf
Incentivized Blockchain-based Social Media Platforms:
A Case Study of Steemit
Chao Li
School of Computing and Information
University of Pittsburgh
Pittsburgh, USA
[email protected]
Balaji Palanisamy
School of Computing and Information
University of Pittsburgh
Pittsburgh, USA
[email protected]
ABSTRACT
Advances in Blockchain and distributed ledger technologies are driving the rise of incentivized social media platforms over Blockchains,
where no single entity can take control of the information and users
can receive cryptocurrency as rewards for creating or curating
high-quality contents. This paper presents an empirical analysis
of Steemit, a key representative of these emerging platforms, to
understand and evaluate the actual level of decentralization and
the practical effects of cryptocurrency-driven reward system in
these modern social media platforms. Similar to Bitcoin, Steemit
is operated by a decentralized community, where 21 members are
periodically elected to cooperatively operate the platform through
the Delegated Proof-of-Stake (DPoS) consensus protocol. Our study
performed on 539 million operations performed by 1.12 million
Steemit users during the period 2016/03 to 2018/08 reveals that the
actual level of decentralization in Steemit is far lower than the ideal
level, indicating that the DPoS consensus protocol may not be a
desirable approach for establishing a highly decentralized social
media platform. In Steemit, users create contents as posts which get
curated based on votes from other users. The platform periodically
issues cryptocurrency as rewards to creators and curators of popular posts. Although such a reward system is originally driven by the
desire to incentivize users to contribute to high-quality contents,
our analysis of the underlying cryptocurrency transfer network
on the blockchain reveals that more than 16% transfers of cryptocurrency in Steemit are sent to curators suspected to be bots and
also finds the existence of an underlying supply network for the
bots, both suggesting a significant misuse of the current reward
system in Steemit. Our study is designed to provide insights on
the current state of this emerging blockchain-based social media
platform including the effectiveness of its design and the operation
of the consensus protocols and the reward system.
ACM Reference Format:
Chao Li and Balaji Palanisamy. 2019. Incentivized Blockchain-based Social
Media Platforms: A Case Study of Steemit. In Proceedings of ACM Conference
on Web Science (WebSci’19). ACM, Boston, MA, USA, 10 pages. https://
doi.org/10.1145/nnnnnnn.nnnnnnn
Permission to make digital or hard copies of all or part of this work for personal or
classroom use is granted without fee provided that copies are not made or distributed
for profit or commercial advantage and that copies bear this notice and the full citation
on the first page. Copyrights for components of this work owned by others than ACM
must be honored. Abstracting with credit is permitted. To copy otherwise, or republish,
to post on servers or to redistribute to lists, requires prior specific permission and/or a
fee. Request permissions from [email protected].
WebSci’19, July 2019, Boston, MA, USA
© 2019 Association for Computing Machinery.
ACM ISBN 978-x-xxxx-xxxx-x/YY/MM. . . $15.00
https://doi.org/10.1145/nnnnnnn.nnnnnnn
1 INTRODUCTION
Recent advances in Blockchain and distributed ledger technologies [27] are driving the rise of incentivized social media platforms
over Blockchains. Examples of such platforms include Steemit1
,
Indorse2
, Sapien3
and SocialX4
. Unique features of these blockchainpowered social media platforms include decentralization of data
generated in the platforms and deep integration of social platforms with the underlying cryptocurrency transfer networks on
the blockchain. Steemit is the first blockchain-powered social media
platform that incentivizes both creator of user-generated content
and content curators. It has kept its leading position during the last
two and a half years and its native cryptocurrency, STEEM, has the
highest market capitalization among all cryptocurrencies issued by
blockchain-based social networking projects. Its market capital is
estimated over 266 million USD in 09/30/2018 [25].
In Steemit, users can create and share contents as blog posts.
Once posted, a blog can get replied, reposted or voted by other
users. Depending on the weight of received votes, posts get ranked
and the top ranked posts make them to the front page. All data
generated by Steemit users are stored in the Steem-blockchain [3].
Similar to other blockchains like Bitcoin [27] and Ethereum [4],
data stored in the Steem-blockchain is publicly accessible and it
is hard to be manipulated. At the core of Steemit are its decentralization and its cryptocurrency-driven reward system used for
rewarding the content creators and curators. Instead of operating
as a single entity like Reddit and Quora, the Steemit platform is
operated by a group of 21 witnesses elected by its shareholders
(uses owning Steemit shares) through the Delegated Proof of Stake
(DPoS) consensus protocol [24]. Unlike traditional social media
platforms that typically do not reward their users, Steemit issues
three types of rewards to its users: (1) producer reward; (2) author
reward; (3) curation reward. The producer rewards are issued to the
elected witnesses producing blocks. It incentivizes users of Steemit
to compete for the top-21 witnesses. The author rewards and curation rewards are issued to users creating posts (authors) and users
voting for posts (curators) respectively. These incentivize authors
to produce posts that attract more votes and curators to vote for
posts that have higher potential to be voted by other users. By the
end of 2018/08, Steemit has issued over 40 million USD worth of
rewards to its users [15].
This paper presents an empirical analysis of Steemit, a key representative of emerging blockchain-based incentivized social media
1https://steemit.com/
2https://indorse.io/
3https://beta.sapien.network/
4https://socialx.network/
arXiv:1904.07310v1 [cs.SI] 15 Apr 2019
WebSci’19, July 2019, Boston, MA, USA Chao Li and Balaji Palanisamy
platforms. Our study targets two core features of Steemit, namely
its decentralized operation and its cryptocurrency-driven reward
system. By analyzing over 539 million operations performed by 1.12
million users during the period 2016/03 to 2018/08, we aim at obtaining several key insights including the answers for the following
set of key questions:
• Do the members of the witness group in the platform have a high
update rate or do the same set of users serve in the group? What is
the power of big shareholders on the decentralization properties
of the platform? Is it possible that a single big shareholder can
determine who to join the witness group?
• What are the factors correlated with rewards issued to authors
and curators? How does the reward system influence users’ behavior? Can incentives be misused by users such as buying votes
from bots to promote their posts?
Our analysis reveals interesting details on the decentralized operation in Steemit and its reward system. Our study on decentralization in Steemit shows that the witness group tends to show a
relatively low update rate and its seats may actually be controlled
by a few large shareholders. Our analysis also indicates that the
majority of top witnesses and top electors form a value-transfer
network. These findings together reveals that the actual level of
decentralization in Steemit is far lower than the ideal level, indicating that DPoS consensus protocol may not be a desirable approach
for establishing a highly decentralized social media platform. Our
analysis of the reward system in Steemit shows that author rewards
earned by an author are correlated with a number of factors, including the number of followers, number of created posts and owned
Steemit shares. However, our analysis of the transfer and vote operations show that more than 16% transfers in the dataset are sent
to curators suspected to be bots. A deeper analysis also reveals the
existence of an underlying supply network for the bots, where big
shareholders delegate their Steemit shares to bots to earn profit.
These results together suggests that the current cryptocurrencydriven reward system in Steemit is under substantial misuse that
deviates from the original intended goal of rewarding high-quality
contents. We point out that the current consensus protocols and
reward systems can hardly achieve their design targets and identify
the key reasons and suggest potential solutions. We believe that
the findings in this paper can facilitate the improvement of existing
blockchain-based social media platforms and the design of future
blockchain-powered websites.
2 BACKGROUND
In this section, we introduce the Steemit social media platform
that runs over the Steem-blockchain[3]. We present the key use
cases of Steemit and discuss how Steemit leverages the underlying
blockchain to function as a decentralized social site that incentivizes
users with cryptocurrency-based rewards.
Steemit. Users of Steemit can create and share contents as blog
posts. A blog post can get replied, reposted or voted by other users.
Based on the weights of received votes, posts get ranked and the
top ranked posts make them to the front page.
Steem-blockchain. Steemit uses the Steem-blockchain[3] to store
the underlying data of the platform as a chain of blocks. Every three
Figure 1: Steem blockchain overview
seconds, a new block is produced, which includes all confirmed operations performed by users during the last three seconds. Steemit
allows its users to perform more than thirty different type of operations. In Figure 1, we display four categories of operations that
are most relevant to the analysis presented in this paper. While
post/vote and follower/following are common features offered by
social sites (e.g., Reddit [31] and Quora [35]), operations such as
witness election and cryptocurrency transfer are features specific
to blockchains.
Witness election and DPoS. Witnesses in Steemit are producers of blocks, who continuously collect data from the entire network, bundle data into blocks and append the blocks to the Steemblockchain. The role of witnesses in Steemit is similar to that of
miners in Bitcoin [27]. In Bitcoin, miners keep solving Proof-ofWork (PoW) problems and winners have the right to produce blocks.
With PoW, Bitcoin achieves a maximum throughput of 7 transactions/sec [7]. However, transaction rates of typical mainstream
social sites are substantially higher. For example, Twitter has an average throughput of more than 5000 tweets/sec [17]. Hence, Steemit
adopts the Delegated Proof of Stake (DPoS) [24] consensus protocol
to increase the speed and scalability of the platform without compromising the decentralized reward system of the blockchain. In
DPoS systems, users vote to elect a number of witnesses as their delegates. In Steemit, each user can vote for at most 30 witnesses. The
top-20 elected witnesses and a seat randomly assigned out of the
top-20 witnesses produce the blocks. With DPoS, consensus only
needs to be reached among the 21-member witness group, rather
than the entire blockchain network like Bitcoin. This significantly
improves the system throughput.
Cryptocurrency - shares and rewards. In Steemit, each vote
casting a post or electing a witness is associated with a weight that
is proportional to the shares of Steemit held by the voter. Like most
blockchains, the Steem-blockchain issues its native cryptocurrencies called STEEM and Steem Dollars (SBD). To receive shares of
Steemit, a user needs to ‘lock’ STEEM/SBD in Steemit to receive
Steem Power (SP) at the rate of 1 ST EEM = 1 SP and each SP
is assigned about 2000 vested shares (VESTS) of Steemit. A user
may withdraw invested STEEM/SBD at any time, but the claimed
fund will be automatically split into thirteen equal portions to be
Incentivized Blockchain-based Social Media Platforms:
A Case Study of Steemit WebSci’19, July 2019, Boston, MA, USA
withdrawn in the next thirteen subsequent weeks. For example, in
day 1, Alice may invest 13 STEEM to Steemit that makes her vote
obtain a weight of 13 SP (about 26000 VESTS). Later, in day 8, Alice
may decide to withdraw her 13 invested STEEM. Here, instead of
seeing her 13 STEEM in wallet immediately, her STEEM balance
will increase by 1 STEEM each week from day 8 and during that
period, her SP will decrease by 1 SP every week.
Through Steem-blockchain, Steemit issues three types of rewards
to its users: (1) producer reward; (2) author reward and (3) curation
reward. The amount of producer reward is about 0.2 STEEM per
block in 2018/08, meaning that a witness producing blocks for a
whole day can earn about 14,400 STEEM. Each day, the Steemblockchain issues a number of STEEM (about 53,800 STEEM per day
in 2018/08) to form the post reward pool and posts compete against
each other to divide up the reward pool based on the total weight
of votes received within seven days from the post creation time.
Here, 75% of reward received by a post goes to the post author and
the rest is shared by the post curators based on their vote weight.
In the rest of this paper, for ease of exposition and comparison, we transfer all values of STEEM/SBD/SP/VESTS to US dollars (USD denoted by $), based on their median transfer rate from
2018/01 to 2018/08, which is $1 = 1 SBD ≈ 0.4 ST EEM = 0.4 SP ≈
800 V ESTS [14].
3 DATASET AND PRELIMINARY RESULTS
In this section, we describe our data collection methodology and
present some preliminary results including the growth of Steemit
over time and the usage of operations in the platform.
3.1 Data collection
The Steem-blockchain offers an Interactive Application Programming Interface (API) for developers and researchers to collect and
parse the blockchain data [13]. From block 1 (created at 2016/03/24
16:05:00) to block 25,563,499 (created at 2018/09/01 00:00:00), we
collected 539,817,204 operations performed by 1,120,166 users of
Steemit during the period 2016/03 to 2018/08. The collected data also
includes 121,619,828 virtual operations performed in the Steemit
platform, such as issuing rewards to witnesses, authors and curators,
during the same time period. In the data collected, we recognized
36 types of operations performed by users and 11 types of virtual
operations. In table 1, we summarize the key operations (OP) and
virtual operations (VO) focused in this paper.
3.2 Preliminary results
Growth of Steemit. We first investigate the growth of Steemit
over time and find that the platform growth is highly impacted by
the cryptocurrency market. In Figure 2, we plot the numbers of
newly registered users and newly performed operations per month
and the changes of Bitcoin and STEEM price during the period
2016/03 to 2018/08. Apart from the initial boost in the first month,
the platform witnessed three times of robust growth, which happened during 2016/05 to 2016/07, 2017/04 to 2017/06 and 2017/11 to
2018/01, respectively. We can observe a strong correlation between
the monthly increment of users and the changes in the STEEM price.
This is due to the fact that more Steemit users investing in trading
may drive up the STEEM price while higher STEEM price may in
OP (social) Description
comment users create posts, reply to posts or replies
vote users vote for posts
custom_json users follow other users, repost a blog
OP (witness-election) Description
witness_update users join the witness pool to be elected, witnesses in pool update their information
witness_vote users vote for witnesses by themselves
witness_proxy users cast votes to the same witnesses voted
by another user by setting that user as their
election proxy
OP (value-transfer) Description
transfer users transfer STEEM/SBD to other users
transfer_to_vesting users transfer STEEM/SBD to VESTS
delegate_vesting_shares users delegate VESTS to other users
withdraw_vesting users transfer VESTS to STEEM
VO (reward) Description
producer_reward platform issues rewards to producers
author_reward platform issues rewards to authors
curation_reward platform issues rewards to curators
Table 1: Summary of operations and virtual operations
turn attract more people joining Steemit. Next, in cryptocurrency
market, the changes of Bitcoin price are usually seen as the most
important market signals. During all the three rising periods, we
see the surge in Bitcoin price, which illustrates the high influence
of the cryptocurrency market on the growth of Steemit and also
suggests that most Steemit users may have a background understanding of cryptocurrency and blockchain. Finally, by comparing
the user curve and the operation curve, we find that even though
the user growth rate drops after all the three rising periods, the
operation growth rate keeps maintaining stability after boosting,
which may reflect that most users who joined Steemit during the
rising periods remain active after the end of the rising periods.
Usage of operations. As discussed in section 3.1, users in Steemit
may perform a variety of operations and these operations are
recorded by the Steem-blockchain. By investigating the usage of
these operations, we aim to answer the following questions: (1)
which categories of operations are more frequently performed by
users?, (2) do users use Steemit more like a social media platform or
like a cryptocurrency wallet? In Figure 3 and Figure 4, we plot the
numbers of social/witness-election operations and value-transfer
operations performed in different months, respectively. The functionality of these operations is described in Table 1. Among the
three categories of operations, the social operations show the highest utilization rate, which indicates that users are using more social
functions offered by Steemit than transfer functions. Among the
three social operations, the vote operation is the most frequently
used one. In 2018/08, more than 21 million votes were cast to posts.
Unlike other voting-based social media platforms such as Reddit, in
Steemit, votes cast by users owning Steemit shares have real monetary value, which may incentivize Steemit users to keep voting
for posts with some frequency to avoid wasting their voting power.
We will discuss more about voting and rewarding mechanisms in
section 5. Among the four value-transfer operations, users perform
the transfer operation more frequently. Since the transfer operation
WebSci’19, July 2019, Boston, MA, USA Chao Li and Balaji Palanisamy
Figure 2: New users/operations per month
(2016/03 to 2018/08)
Figure 3: New social/witness-election operations per month (2016/03 to 2018/08)
Figure 4: New value-transfer operations per
month (2016/03 to 2018/08)
is the only operation among the four that is not associated with
VESTS, namely shares of Steemit, the fact may reflect that more
trading behaviors are happening in Steemit than investing behaviors. Finally, the number of performed witness-election operations
are relatively small, compared with the other two categories. Each
month, we see only thousands of users setting or updating their witness votes through the witness_vote and witness_proxy operations
and only hundreds of users joining the witness pool or updating
their witness information through the witness_update operation.
This reflects a relatively low participation rate in witness election
regarding both elector and electee and could impact the level of
decentralization in the platform. We discuss witness election in
more detail in section 4.
4 DECENTRALIZATION IN STEEMIT
As a blockchain-based social media platform, Steemit distinguishes
itself from traditional social sites through the decentralization
brought by the blockchain. In this section, we investigate the actual level of decentralization in Steemit by analyzing the witness
election process of DPoS in detail.
4.1 Decentralized platform operation
Centralization and decentralization. In traditional social media
platforms, such as Reddit and Quora, a single entity (i.e., Reddit, Inc.
and Quora, Inc.) owns the complete data generated by users and
operates the websites. In contrast, Steemit not only open sources
its front-end, Condenser and the back-end Steem-blockchain [30],
but also makes all its data in the blockchain available for public
access [3]. Rather than functioning as a single entity, the Steemit
platform is operated by a group of 21 witnesses elected by its users.
Any user in Steemit can run a server, install the Steem-blockchain
and synchronize the blockchain data to the latest block. Then, by
sending a witness_update operation to the network, the user can
become a witness and have a chance to operate the website and
earn producer rewards if he or she can gather enough support from
the electors to join the 21-member witness group.
Steemit-2 and STEEM-2. As anyone can copy the code and data
of Steemit, one may naturally doubt that an adversary, say Bob, may
build a ‘fake’ Steemit platform, Steemit-2 that has exactly same the
functionality and historical data as Steemit. To distinguish Steemit-2
from Steemit, we name the cryptocurrency issued by Steemit-2 as
STEEM-2 (and also SBD-2). Here, a natural question that arises is
what makes people believe that Steemit, rather than Steemit-2, is
the ‘real’ one? In the decentralized network, the opinion of ‘which
one is real’ is determined by the consensus of Steemit users or, to be
more precise, the shareholders. With the DPoS consensus protocol,
each block storing data of Steemit is signed by a top witness elected
by the shareholders, which may represent the consensus among
the shareholders. Therefore, unless most of shareholders switch to
Steemit-2, the new blocks generated in Steemit-2 will be signed by
witnesses elected by a few shareholders and will not be recognized
by the entire community.
Factors affecting decentralization of Steemit. In our work, we
study the characteristics of decentralization in Steemit by analyzing
the witness election process. In general, we consider Steemit to
have an ideal level of decentralization if members of the 21-member
witness group are frequently updated and if these members all have
different interests. We also consider Steemit to have a relatively high
decentralization closer to the ideal level if it allows more people to
join the 21-member witness group, if the power of big shareholders
is not decisive and if the election is not heavily correlated with valuetransfer operations. We investigate these aspects in the following
subsection.
4.2 Analyzing witness election
Update of the 21-member witness group. To investigate the
update of the 21-member witness group, we extract the producer
of each block from block 1 to block 25,563,499 and plot the results
as a heatmap in Figure 5. We first compute the number of blocks
produced by each witness and sort the witnesses based on their
produced blocks in total. For the top-30 witnesses that have the
highest number of produced blocks, we plot their attendance rate
in the 21-member witness group during the thirty months. During
a month that has thirty days, there should be 30 ∗ 24 ∗ 60 ∗ 60/3 =
864,000 blocks generated because blocks are generated every three
seconds. For every 21 blocks (63 seconds), the 21 elected witnesses
are shuffled to determine their order for generating the next 21
blocks. Therefore, if a witness has a 100% attendance rate in the
elected group, it can at most produce 864000/21 ≈ 41,142 blocks
in a thirty-day month. In Figure 5, we find that most of the top-30
witnesses, once entering the 21-member witness group, maintained
a high attendance rate closer to 100% for a long time. From month 12
to month 30, namely one and a half year, the top-12 witnesses firmly
held at least 10 seats, nearly half of positions in the 21-member
Incentivized Blockchain-based Social Media Platforms:
A Case Study of Steemit WebSci’19, July 2019, Boston, MA, USA
Figure 5: Heatmap of top-30 witnesses’ attendance rate in the 21-
member witness group during 30 months from 2016/03 to 2018/08
witness group. From month 23 to month 30, 17 seats were held
by 18 witnesses. From witness 13 to witness 30, we can observe
a transition period, namely month 15 to month 20, during which
the places of nine old witnesses were gradually taken by nine new
witnesses. Overall, the 21-member witness group tends to show
a relatively low update rate. The majority of seats were firmly
controlled by a small group of witnesses. We do observe some
switch of seats but that happened only in a low frequency.
Power of big shareholders. Next, we investigate the influence
of big shareholders in the witness election. As described in Table 1,
a user has two ways to vote for witnesses. The first option is to
perform witness_vote operations to directly vote for at most 30 witnesses. The second option is to perform a witness_proxy operation
to set another user as an election proxy. For example, Alice may set
Bob to be her proxy. Then, if both Alice and Bob own $100 worth of
shares, any vote cast by Bob will be associated with a weight of $200
worth of shares. Once Alice deletes the proxy, the weight of Bob’s
votes will reduce to $100 worth of shares immediately. In Figure 6,
we plot the stacked bar chart representing a snapshot of weighted
votes received by the top-60 witnesses, who have produced the
highest amounts of blocks, at block 25,563,499. The figure shows
the distribution of votes cast by the top-29 electors whose votes
have the highest weight, either brought by their own shares or
shares belonging to users setting these electors as proxy. The sum
of weighted votes cast by all other electors outside the top-29 is
represented as ‘sum of rest’. From Figure 6, we see a few top electors
have their votes weighted by striking amounts of shares. The top-1
elector has his or her votes weighted by $19,800,000 worth of shares.
A deeper investigation regarding the top-1 elector shows that all
the shares affecting his or her weight are not directly owned by
this user, but owned by another user, who set the top-1 elector as
proxy. The runner-up elector, which is the account belonging to the
main exchange used by Steemit users, has its votes associated with
a weight of $12,100,000 worth of shares. From the figure, we see
all the 27 witnesses voted by the top-1 elector enter the top-50, all
the 18 witnesses receiving votes from both the top-2 electors enter
the top-30 and the only witness receiving votes from all the top-3
Figure 6: A snapshot of weighted votes received by top-60 witnesses at block 25,563,499. The votes are weighted by shares (estimated in USD) owned by the electors. We show the votes cast by the
top-29 electors owning the highest weight and use ‘sum of rest’ to
represent the sum of weighted votes cast by all other electors.
electors becomes the top-1 witness. We can also observe that 19 out
of the top-20 witnesses receive at least two votes from the top-3
electors and 29 out of the top-30 witnesses receive at least one vote
from the top-3 electors. As illustrated by the results, the distribution of weight of votes in witness election is heavily skewed, which
suggests that the election of 21 witnesses may be significantly impacted by a few big shareholders, This phenomenon may not be a
good indication for a decentralized social media platform. In the
worst case, if the 21-member witness group is controlled by a single shareholder, the platform will simply function as a centralized
model.
Value-transfer operations among election stakeholders. Finally, we investigate the value-transfer operations performed among
the top-30 witnesses, top-29 electors and the accounts selecting
top-29 electors as proxy. The data is plotted as a directed graph in
Figure 7, where edges are colored by their source node color. The
edge thickness represents the amount of transferred value from
source to target, which is the sum of value transferred through transfer and transfer_to_vesting operations. Since most Steemit users use
runner-up elector to trade cryptocurrency, the graph does not show
edges connected with that account. Our first observation about
this graph is that only two out of top-30 witnesses and three out
of top-29 electors never perform value-transfer operations while
all other investigated users form a value-transfer network, which
has 3.34 average degree, 0.21 average clustering coefficient and 3.96
average path length. In the network, we find a cluster of users who
select top-29 electors as proxy. After manually checking the profiles
of these users, we find that this cluster represents a community of
Korean users, which is connected to the rest of the network mainly
through several leaders of the Korean community. Overall, what
we observed from the value-transfer operations suggests that the
majority of the investigated election stakeholders have economic
interactions, which may not be a good indication of a perfectly
decentralized witness group where the members are expected to
have different interests.
WebSci’19, July 2019, Boston, MA, USA Chao Li and Balaji Palanisamy
Figure 7: Graph of value-transfer operations performed among
the top-30 witnesses, the top-29 electors and the accounts selecting
these electors as proxy. The graph was plotted using Gephi [2].
4.3 Discussion
Our study in this section demonstrates that the 21-member witness
group tends to show a relatively low update rate and these seats
may actually be controlled by a few big shareholders. Our study
also indicates that the majority of top witnesses and top electors
form a value-transfer network and have economic interactions. Together, these results suggest that the actual level of decentralization
in Steemit is far lower than the ideal level. One key reason for the
low decentralization is the use of DPoS consensus protocol. The
DPoS consensus protocol has been widely adopted by mainstream
blockchain-based platforms such as BTS and EOS and has been
proved to be an effective approach of enhancing transaction rates
of blockchains. However, there has been a long debate surrounding decentralization of DPoS. The opponents of DPoS censure that
DPoS-powered platforms trade decentralization for scalability as
the consensus in these platforms is only reached among a small
committee (e.g., the 21-member witness group in Steemit), instead
of among all interested members (e.g., all miners in Bitcoin and
Ethereum powered by Proof-of-Work (PoW) consensus). The supporters of DPoS argue that those PoW-powered platforms has been
controlled by a few mining pools, showing even lower decentralization than the DPoS-powered platforms. The data-driven analysis
in this section deeply investigates the underlying behaviors of participants interested in the core component of DPoS, namely the
witness election. The results reveal that the current electoral system is making the decentralization quite fragile, as the committee
intended to be decentralized is actually quite centralized in practice.
A quick solution to address the symptoms is to restrict the power of
big stakeholders, such as cutting the number of times that big stakeholders can vote in the election. A better way of solving the problem
is to replace DPoS with more advanced consensus protocols, that
can form the committee without an election involving interactions
among users. For example, Algorand [10], a recent cryptocurrency,
proposed a new Byzantine Agreement (BA) protocol that makes the
election be fairly performed by the Verifiable Random Functions in
cryptography, rather than by users.
5 REWARD SYSTEM IN STEEMIT
A core feature of Steemit is its reward system. As a blockchainbased social media network that issues its own cryptocurrencies,
Steemit leverages its native cryptocurrencies to incentivize authors
and curators of posts for rewarding contributors to the platform,
who either create good contents or screen out good contents. In this
subsection, we theoretically analyze the reward system in Steemit
and study the factors correlated with rewards earned by authors and
curators. We also jointly investigate the value-transfer operations
and vote operations to understand if the reward system is being
misused by Steemit users, such as buying votes from bots to promote
their posts for the purpose of earning higher profit. Such misuse
may deviate from the original intended goal of the Steemit reward
system.
5.1 Reward system
To the best of our knowledge, Steemit [16] never formally published
the details of its reward system, so we investigated its source code
to understand its reward system [30] in detail. Each time a user
votes for a post, the vote contributes a certain amount of reward
shares (rshares, denoted as rs) to that post. It is computed as
rs = e_V ESTS ∗
vp ∗ vw − 0.0049
50
where e_V ESTS refers to effective vesting shares (VESTS),vp stands
for voting power and vw denotes voting weight. In Steemit, users
can deposit or withdraw VESTS through transfer_to_vesting operation and withdraw_vesting operation, respectively. However,
as described in Table 1, a user may also delegate VESTS to other
users through delegate_vesting_shares operation. For example, if
Alice owns 100 VESTS and if she delegates 50 VESTS to Bob and
if she also receives 30 VESTS delegated to her by Carol, then her
e_V ESTS = 100 − 50 + 30 = 80 V ESTS. A user may set voting
weight vw to any value between 0% and 100%. Steemit leverages
voting power vp to restrict the number of weighted votes cast by
users per day. Initially, each user has vp = 100%. Then, if a user
casts a vote to a post, this user’s vp will drop to (1−
1−0.0049
50 )∗100%,
which is roughly 98% if he/she sets vw = 100%. That is, if a user
keeps voting, his/her vp will also keep dropping. Once vp drops
to 0%, his/her vote will contribute no rshares to posts and the user
and post authors will not earn rewards from this vote. Each day, vp
recovers 20% before it is back to 100%.
A post, after being created, can accumulate rshares from received
votes during a 7-day time window. At the end of the time window,
the post will use the accumulated rshares to compete with accumulated rshares of other posts to divide up the post reward pool
(about 53,800 STEEM per day in 2018/08). After that, 75% of the
post reward (denoted as pr) is directly issued to the post author as
author reward (denoted as ar) and the rest 25% is finally shared by
all curators who voted for the post during the 7-day time window,
namely their curation rewards (denoted ascr). The curation reward
received by a single curator can be computed as:
cr = 0.25 ∗ pr ∗
√
rsb + rs −
√
rsb
√
rsT
∗ min(
td
30
, 1)
As can be seen,cr is computed from 25% of pr, but is affected by two
factors. The first factor is the ratio between √
rsb + rs −
√
rsb and
Incentivized Blockchain-based Social Media Platforms:
A Case Study of Steemit WebSci’19, July 2019, Boston, MA, USA
(a) Followers (b) Post numbers (c) Vesting shares
Figure 8: Scatter plot investigating correlation between followers (post numbers, vesting shares) and author rewards
Figure 9: Distribution of accumulated
rshares over time after post creation time
Figure 10: Distribution of accumulated curation rewards over time
Figure 11: CDF investigating the time gap between post creation time and bot voting time
√
rsT , where rsb denotes the rshares accumulated by the post before
this user’s vote and rsT refers to the total accumulated rshares by
the post during the 7-day time window. This is a very interesting
factor as it suggests that curators who want to earn higher cr
should vote for posts as early as possible to make rsb
smaller. It
also suggests that such curators vote for posts that have higher
probability to accumulate higher rsT by the end of 7-day time
window. However, the second factor forces curators to reconsider
the early vote strategy. It is the minimum of td
30 and 1, where td
denotes the time difference in minutes between the post creation
time and voting time. For example, if Alice votes for a post one
minute after the post creation time, she will only receive 1
30 of the
curation rewards assigned to her while the rest 29
30 is again issued to
the post author. Instead, if Alice casts her vote 30 minutes after the
post creation time, she can completely earn the assigned curation
rewards. Because of the second factor, a post author usually receives
more than 75% of the post reward. In summary, the two factors in
the curation reward equation make it hard to determine the best
strategy of the voting time in the reward earning game. However,
it is clear that voting for posts that are likely to be voted by more
users in the future is certainly helpful.
5.2 Factors correlated with rewards
After theoretically analyzing the reward system, we now investigate
the blockchain data to learn the factors correlated with author
rewards and curation rewards.
Author rewards. Regarding author rewards, we investigate three
factors: (1) number of followers owned by an author; (2) number of
posts created by an author; (3) amount of vesting shares (VESTS)
owned by an author. In order to gauge the correlation between
these factors and author rewards, for each factor, we plot in Figure 8
the average and median of values of the factor (y) against author
rewards estimated in USD (x). Following the methodology used by
Kwak et al. [23], we also bin the value of author rewards in log scale
and plot the average and median per bin in lines. In Figure 8(a), we
see the line of average is always above the line of median, indicating
that some users with followers more than average do not receive
expected degree of support from their followers regarding their
posts. The gap between the two line is larger for users receiving
lower author rewards, but becomes relatively small for the top
authors earning rewards more than $1000. The median number of
followers grows steadily without showing a flat period or a surge.
Next, in Figure 8(b), we investigate correlation between number
of posts created by an author and rewards received by an author.
Again, we find that the line of average keeps itself above the line of
median, indicating that some users post more blogs than average but
do not receive more rewards as they may expect. This may suggest
that some authors slow down their posting speed while improving
the quality of their posts. Until $30,000 rewards, the median number
of posts shows a relatively stable growth. After that, it experiences
a fluctuation among the top authors. Finally, in Figure 8(c), we
investigate correlation between VESTS owned by an author and
WebSci’19, July 2019, Boston, MA, USA Chao Li and Balaji Palanisamy
rewards received by an author. From the observation that the line of
median lies nearly always below the line of average, we may infer
that some users own higher shares of Steemit than average but fail
to increase their author rewards to the same scale. By observing
the line of median, we find that the median number of VESTS
stays relatively flat at $12.5 before author rewards reach $10. The
$12.5 worth of VESTS is equal to 5 STEEM, which is the minimum
line that Steemit suggests that its user transfer to VESTS for the
purpose of maintaining the user’s account in a normal state. From
$10 worth of author rewards, the median points start to disperse.
It is interesting to find that many authors earning rewards more
than $10 maintain very low VESTS, even lower than the suggested
minimum line. However, beyond $10 worth of author rewards, the
line of median still shows a positive trend.
Overall, our results illustrate a positive correlation between all
the three factors and author rewards. Our results do not imply
causation, but they suggest that users, both authors and curators,
consider the three factors as indicators of the potential popularity
of posts.
Curation rewards. As we have analyzed in section 5.1, the curation reward assigned to a single curator is affected by two factors.
The first factor suggests that reward-driven curators cast their votes
early and vote for posts with high potential for popularity. However, the second factor penalizes the early votes. Regarding the
potential popularity of posts, a reward-driven curator may leverage
the number of followers/posts/VESTS of post authors as indicators.
However, due to the conflict suggestions from the two factors in
the curation reward computation equation, it is hard to determine
the best voting time for maximizing curation rewards. Therefore,
we decide to investigate what are the choices made by users in the
data set. To observe decisions made by different levels of shareholders, we divide users into four groups based on their owned VESTS,
namely low (< 104
), medium (104 − 106
), high (106 − 108
) and top
(> 108
). In Figure 9, we plot the distribution of the accumulated
rshares that have been contributed by all votes to all posts every
second after the post creation time. Since rshares is the most important factor of post rewards and also the key value of votes, this
figure helps revealing users’ strategies on selecting voting time. As
can be seen from results, there are two peaks that happened. The
first peak is at the sixth second, which indicates that many curators
choose to vote for a post after six seconds (two blocks) of the post
creation time. This may reflect that many users are still using the
early vote strategy. The second peak is more interesting. It happens
at the 1803 second, namely the first block after the 30-minute time
period penalized by the second factor of the curation reward computation equation. This reflects that some users do understand the
punishment mechanism and are deliberately avoiding from being
penalized. We can also observe that the low-level shareholders tend
to not pay special attention to the voting strategies while all the
other three groups of shareholders tend to contribute more to the
two peaks and are more likely to vote earlier besides the two peaks.
The results of the distribution of accumulated curation rewards
along time are shown in Figure 10, which show a clear effect of the
punishment mechanism on the early votes. The results suggest that
curators who want to earn more rewards vote around the time of
the second peak, namely the end of the 30-minute penalty period.
5.3 Misuse of reward system
The reward system in Steemit may be a good way to reward contributors to the platform, but it may also be misused by some users
in ways that deviate from the original intended goal in Steemit,
such as buying votes from bots to promote some meaningless posts
for the purpose of earning profit. In this subsection, we aim at
understanding to what extent such behaviors are performed in
Steemit.
Temporal correlation. We study this topic by investigating the
temporal correlation between transfer operations and vote operations. Concretely, if there is a suspicious vote cast by a curator to an
author through a vote operation, there should also be a suspicious
fund transferred to the curator from the author through a transfer
operation that happens before the vote operation is performed by
the curator. Specifically, we consider a transfer operation is suspicious if the ‘memo’ area (allowing sender to leave a message) of the
transfer operation only contains a link pointing to a recent post
created by the sender. If in addition, the recipient of the transfer
operation, after receiving the fund from the sender, votes for the
post matching that link within the 7-days time window after the
post creation time, we consider it as a suspicious trade between a
post author and a voting bot.
After parsing the 18,172,530 transfer operations in blockchain, we
find that 5,031,737 (27.69%) of them only contain a single post link in
the ‘memo’ area. By further investigating the vote operation, we find
that 2,939,051 (16.17%) of the total transfer operations are followed
by a suspicious vote operation voting for the post link within the
7-days time window, which are performed by the recipients of the
funds. Among all the fund recipients, the top one has performed
such trade for 113,068 times and earned nearly 1 million USD.
Voting time of suspicious curators. To further investigate the
features of the curators suspected to be bots, in Figure 11, we plot
the cumulative distribution function (CDF) of the time gap between
the post creation time and voting time. The solid line refers to the
CDF of votes cast by all curators who ever received any amount
of fund from the authors of voted posts at any time point in the
past. The dotted line describes the CDF of votes cast by curators
suspected to be bots. As can be seen, the curators suspected to be
bots cast their votes much earlier than the average. The dotted line
reaches 98.9% at the 604,800 second, which refers that nearly all
their votes are sent within the 7-day time window and they are
internationally contributing rshares of their votes to the posts.
An example bot network. In Figure 12, we plot an example
bot network, which describes the transfer operations sent to the
curators suspected to be bots from the top-30 post authors, who
most frequently had contacted with these suspected curators. The
node sizes are weighted by weighted in-degree. As can be seen,
the thirty authors contacted a large number of suspected curators
because they may need rshares from more than one suspected
curator to promote their posts. We can also observe that most
suspected curators have their account name containing suspected
keywords, such as boost, promote, whale and upvote.
An example supply network for bots. Finally, we analyze where
do these suspected curators get large amounts of VESTS to make
their votes weighted by higher rshares. After a deep analysis, we
find that there is an underlying supply network, where some big
Incentivized Blockchain-based Social Media Platforms:
A Case Study of Steemit WebSci’19, July 2019, Boston, MA, USA
Figure 12: Graph of transfer operations sent from selected post
authors (red) to the curators suspected to be bots (blue). The graph
was plotted using Gephi [2].
Figure 13: Graph of delegate_vesting_shares operations and transfer operations among the top-30 curators suspected to be bots (red)
and their suspected suppliers (blue). The graph was plotted using
Gephi [2].
shareholders delegate their VESTS to the suspected curators through
delegate_vesting_shares operations and the suspected curators periodically pay ‘rent’ to these VESTS suppliers through transfer operations. In Figure 13, we plot an example supply network for bots,
which describes the delegate_vesting_shares operations sent to the
top-30 suspected curators with highest earnings and the transfer
operations sent from the top-30 suspected curators with amount
higher than $100. The node size reflects weighted degree and the
edge thickness is weighted by the amount of transferred value.
From the graph, we find that most of the top-30 suspected curators
own a number of suppliers and a few of them have many. It is
interesting to find that the top-1 supplier is the same user who also
stands behind the top-1 witness presented in section 4.2. We find
that this user has earned more than 1.5 million USD by delegating
his or her VESTS to the curators suspected to be bots.
5.4 Discussion
In summary, we have discussed the design principle, current use
status and underlying misuse of the cryptocurrency-driven reward
system of Steemit in this section. The rise of social bots and the
harm caused by them to the online ecosystems has been widely recognized [9]. Our results in this section reveal that bots have broadly
appeared in the emerging incentivized blockchain-based social media networks. In Steemit, the monetary value of cryptocurrencies
and the deep integration of social operations and value-transfer operations have jointly led to the special supply chain, where bots buy
voting powers from big stakeholders and sell the voting powers to
users who want to promote posts. Clearly, such behaviours violate
the original intention of the reward system and may eventually result in a front page occupied by valueless posts or spams promoted
by bots. Moreover, due to the decentralized operation, it is difficult
to delete these problem posts appeared at the front page because
no single entity has this power. Our discussion in this section also
suggests a way of detecting bots through investigating the temporal
correlation between transfer operations and vote operations. We believe this detection approach is effective at the current stage when
bots prefer to trade with the native cryptocurrency ST EEM, which
is detectable through the Steem-blockchain. However, smarter bots
may hide the temporal correlation between transfer operations and
vote operations by asking their clients to pay them through other
cryptocurrencies such as Bitcoin or even privacy-preserving cryptocurrencies such as Zerocash [28]. Then we need to rely on other
approaches, such as Deep Neural Networks [22], to detect bots.
6 RELATED WORK
In recent years, due to the rapid growth and consistent popularity, social media platforms have received significant attention
from researchers. A large number of research papers have analyzed several popular social media platforms from various perspectives [1, 8, 11, 12, 29, 31, 32, 35, 36]. Tan et al. [32] investigated
users’ behavior in Reddit and found that users continually post
in new communities. Singer et al. [29] observed a general quality
drop of comments made by users during activity sessions. Hessel et al. [12] investigated the interactions between highly related
communities and found that users engaged in a newer community
tend to be more active in their original community. In [11], the
authors studied the browsing and voting behavior of Reddit users
and found that most users do not read the article that they vote
on. An extensive survey of recent research on Reddit is provided
in [26]. Besides Reddit, several other social media platforms have
also been analyzed by researchers. Wang et al. [35] analyzed the
Quora platform and found that the quality of Quora’s knowledge
base is mainly contributed by its user heterogeneity and question
graphs. Anderson et al. [1] investigated the Stack Overflow platform and observed significant assortativity in the reputations of
co-answerers, relationships between reputation and answer speed.
Recent advances in blockchain and distributed ledger technologies in terms of scalability [18, 20], efficiency [6, 34] and privacy [5,
21, 28] have empowered blockchains to support various services
WebSci’19, July 2019, Boston, MA, USA Chao Li and Balaji Palanisamy
beyond money transfer, including incentivized blockchain-based
social media platforms such as Steemit. Recently, this new type of
social media platform has drawn some attention from researchers.
Thelwall et al. [33] analyzed the first posts made by 925,092 Steemit
users to understand the factors that may drive the post authors
higher rewards. Their results suggest that new users of Steemit
start from a friendly introduction about themselves rather than immediately providing useful content. In a very recent work, Kiayias
et al. [19] studied the decentralized content curation mechanism
from a computational perspective. They defined an abstract model
of a post-voting system, along with a particularization inspired by
Steemit. Through simulation of voting procedure under various conditions, their work identified the conditions under which Steemit
can successfully curate arbitrary lists of posts and also revealed
the fact that selfish participant behavior may hurt curation quality.
To the best of our knowledge, the work presented in this paper is
the first research that investigates the reward system and decentralization features of incentivized blockchain-based social media
platforms through a rigorous empirical analysis of the operations
reflected in the underlying blockchain data.
7 CONCLUSION
In this paper, we presented an empirical analysis of Steemit, a
blockchain-based incentivized social media platform where no single entity can take control of the information and users are rewarded
for the contributions they make. We analyzed over 539 million operations performed by 1.12 million users during the period 2016/03
to 2018/08. Our results show interesting details about two core
features of Steemit, namely its decentralized management and its
reward system. Our study on decentralization in Steemit shows
the actual level of decentralization in Steemit is far lower than the
ideal level, indicating that DPoS consensus protocol may not be a
desirable approach for establishing a highly decentralized social
media platform. Our analysis of the reward system reveals the fact
that more than 16% transfers of cryptocurrency in Steemit are sent
to curators suspected to be bots and also finds the existence of an
underlying supply network for the bots, both suggesting that the
current cryptocurrency-driven reward system in Steemit is under
severe misuse that deviates from the original intended goal of rewarding high-quality contents.. Overall, we believe that the results
in this paper provide insights on the current state of the emerging
blockchain-based social media platforms including the effectiveness
of the design and the operation of the consensus protocols and the
reward system.
REFERENCE
[1] Ashton Anderson et al. Discovering value from community activity on focused
question answering sites: a case study of stack overflow. In Proceedings of the
18th ACM SIGKDD, pages 850–858. ACM, 2012.
[2] Mathieu Bastian et al. Gephi: an open source software for exploring and manipulating networks. Icwsm, 8(2009):361–362, 2009.
[3] Steem blockchain [Internet]. Available from. https://developers.steem.io/. Accessed Oct. 2018.
[4] Vitalik Buterin et al. A next-generation smart contract and decentralized application platform. white paper, 2014.
[5] Jan Camenisch, Manu Drijvers, and Maria Dubovitskaya. Practical uc-secure
delegatable credentials with attributes and their application to blockchain. In
Proceedings of the 2017 CCS, pages 683–699. ACM, 2017.
[6] Melissa Chase and Sarah Meiklejohn. Transparency overlays and applications. In
Proceedings of the 2016 ACM SIGSAC Conference on Computer and Communications
Security, pages 168–179. ACM, 2016.
[7] Kyle Croman et al. On scaling decentralized blockchains. In International Conference on Financial Cryptography and Data Security, pages 106–125, 2016.
[8] Munmun De Choudhury and Sushovan De. Mental health discourse on reddit:
Self-disclosure, social support, and anonymity. In ICWSM, 2014.
[9] Emilio Ferrara, Onur Varol, Clayton Davis, Filippo Menczer, and Alessandro
Flammini. The rise of social bots. Communications of the ACM, 59(7):96–104,
[10] Yossi Gilad, Rotem Hemo, Silvio Micali, Georgios Vlachos, and Nickolai Zeldovich.
Algorand: Scaling byzantine agreements for cryptocurrencies. In Proceedings of
the 26th Symposium on Operating Systems Principles, pages 51–68. ACM, 2017.
[11] Maria Glenski, Corey Pennycuff, and Tim Weninger. Consumers and curators:
Browsing and voting patterns on reddit. IEEE Transactions on Computational
Social Systems, 4(4):196–206, 2017.
[12] Jack Hessel et al. Science, askscience, and badscience: On the coexistence of
highly related communities. In ICWSM, pages 171–180, 2016.
[13] Interactive Steem API [Internet]. Available from. https://steem.esteem.ws/.
Accessed Oct. 2018.
[14] STEEM Price [Internet]. Available from. https://coinmarketcap.com/currencies/
steem/. Accessed Oct. 2018.
[15] Steem Total Rewards [Internet]. Available from. https://steem.io/. Accessed Oct.
[16] Steem Whitepaper [Internet]. Available from. https://steem.io/steem-whitepaper.
pdf. Accessed Oct. 2018.
[17] Twitter Usage Statistics [Internet]. Available from. http://www.internetlivestats.
com/twitter-statistics/. Accessed Oct. 2018.
[18] Harry Kalodner et al. Arbitrum: Scalable, private smart contracts. In Proceedings
of the 27th USENIX Conference on Security Symposium, pages 1353–1370. USENIX
Association, 2018.
[19] Aggelos Kiayias, Benjamin Livshits, Andrés Monteoliva Mosteiro, and Orfeas
Stefanos Thyfronitis Litos. A puff of steem: Security analysis of decentralized
content curation. arXiv preprint arXiv:1810.01719, 2018.
[20] Eleftherios Kokoris-Kogias et al. Omniledger: A secure, scale-out, decentralized
ledger via sharding. In 2018 IEEE Symposium on Security and Privacy (SP), pages
583–598. IEEE, 2018.
[21] Ahmed Kosba et al. Hawk: The blockchain model of cryptography and privacypreserving smart contracts. In 2016 IEEE symposium on security and privacy (SP),
pages 839–858. IEEE, 2016.
[22] Sneha Kudugunta and Emilio Ferrara. Deep neural networks for bot detection.
Information Sciences, 467:312–322, 2018.
[23] Haewoon Kwak, Changhyun Lee, Hosung Park, and Sue Moon. What is twitter, a
social network or a news media? In Proceedings of the 19th international conference
on World wide web, pages 591–600. ACM, 2010.
[24] Daniel Larimer. Delegated proof-of-stake (dpos). Bitshare whitepaper, 2014.
[25] Steem market cap [Internet]. Available from. https://coinmarketcap.com/
currencies/steem/. Accessed Oct. 2018.
[26] Alexey N Medvedev et al. The anatomy of reddit: An overview of academic
research. arXiv preprint arXiv:1810.10881, 2018.
[27] Satoshi Nakamoto. Bitcoin: A peer-to-peer electronic cash system. 2008.
[28] Eli Ben Sasson et al. Zerocash: Decentralized anonymous payments from bitcoin.
In 2014 IEEE Symposium on Security and Privacy (SP), pages 459–474. IEEE, 2014.
[29] Philipp Singer, Emilio Ferrara, Farshad Kooti, Markus Strohmaier, and Kristina
Lerman. Evidence of online performance deterioration in user sessions on reddit.
PloS one, 11(8):e0161636, 2016.
[30] Steemit souce code [Internet]. Available from. https://github.com/steemit. Accessed Oct. 2018.
[31] Greg Stoddard. Popularity and quality in social news aggregators: A study of
reddit and hacker news. In Proceedings of the 24th WWW, pages 815–818. ACM,
[32] Chenhao Tan and Lillian Lee. All who wander: On the prevalence and characteristics of multi-community engagement. In Proceedings of the 24th WWW, pages
1056–1066, 2015.
[33] Mike Thelwall. Can social news websites pay for content and curation?
the steemit cryptocurrency model. Journal of Information Science, page
0165551517748290, 2017.
[34] Alin Tomescu and Srinivas Devadas. Catena: Efficient non-equivocation via
bitcoin. In 2017 38th IEEE Symposium on Security and Privacy (SP), pages 393–409.
IEEE, 2017.
[35] Gang Wang, Konark Gill, Manish Mohanlal, Haitao Zheng, and Ben Y Zhao.
Wisdom in the social crowd: an analysis of quora. In Proceedings of the 22nd
international conference on World Wide Web, pages 1341–1352. ACM, 2013.
[36] Diyi Yang and Scott Counts. Understanding self-narration of personally experienced racism on reddit. In ICWSM, pages 704–707, 2018.