CoChat
Sign Up
Cited by 91Semantic Web and OntologiesResearch Data Management PracticesScientific Computing and Data Management

The DOI matches its Crossref record; 40 of the 42 references it cites were found in a registry or an index; none of the 13 sources its claims cite is checked for support yet.

Decentralized provenance-aware publishing with nanopublications

Open access
Kuhn, Tobias, Chichester, Christine, Krauthammer, Michael, Queralt-Rosinach, Núria, Verborgh, Ruben, Giannakopoulos, George, Ngonga Ngomo, Axel-Cyrille, Viglianti, Raffaele, Dumontier, Michel·2016·PeerJ Computer Science
DOI 10.7717/peerj-cs.78

Published version of a preprint

Open this paper in CoChat — free account

Abstract

Publication and archival of scientific results is still commonly considered the responsability of classical publishing companies. Classical forms of publishing, however, which center around printed narrative articles, no longer seem well-suited in the digital age. In particular, there exist currently no efficient, reliable, and agreed-upon methods for publishing scientific datasets, which have become increasingly important for science. In this article, we propose to design scientific data publishing as a web-based bottom-up process, without top-down control of central authorities such as publishing companies. Based on a novel combination of existing concepts and technologies, we present a server network to decentrally store and archive data in the form of nanopublications, an RDF-based format to represent scientific data. We show how this approach allows researchers to publish, retrieve, verify, and recombine datasets of nanopublications in a reliable and trustworthy manner, and we argue that this architecture could be used as a low-level data publication layer to serve the Semantic Web in general. Our evaluation of the current network shows that this system is efficient and reliable.

Overview

AI snapshot

Written from the paper’s full text.

TL;DR

The paper proposes a decentralized, provenance-aware publishing architecture for scientific data as nanopublications with trusty URIs, distributed across a peer-to-peer server network. Evaluation of a small server network shows efficient replication and reliable retrieval, including under simulated unreliable connections.

Key findings

  • In the first evaluation, five existing nanopublication datasets (>5 million nanopublications, close to 200 million RDF triples, 15.6 GB uncompressed TriG) were replicated across three servers in 13 h 21 min, averaging about 400,000 new nanopublications/hour (>100/s); 4,802 monitor HTTP requests had no timeouts, errors, or unverifiable nanopublications, with response times <0.21 s on the same continent, 99.77% below 0.5 s, and always below 1.1 s.
  • Under stress test, the nanopublication server handled 113,178 requests (average 377 requests/s) with average response time 0.12 s; the SPARQL endpoint needed 100 times longer (13 s average), handled 1,267 requests, and hit the 60 s timeout when more than 40 clients accessed it in parallel.
  • Client-side retrieval of the LIDDI dataset (~100,000 triples) succeeded in all 40 attempts; normal connection averaged 235 s (max 279 s), simulated unreliable connection averaged 517 s (max below 10 min), and all normalized downloads were identical.
  • Overall replication capacity was estimated at around 9.4 million new nanopublications/day or 3.4 billion/year; the network later consisted of 15 server instances on 10 sites in 8 countries.

Methods

  • Design: a multi-layer architecture with a low-level decentralized nanopublication server network using trusty URIs, REST API, journals, surface patterns, and nanopublication indexes; servers replicate by polling peers.
  • Data: seven existing large nanopublication datasets are listed; the first evaluation used five of them (>5 million nanopublications, close to 200 million triples, 15.6 GB as uncompressed TriG), and the third used LIDDI (~100,000 triples).
  • Evaluation: three parts—(1) load five datasets onto three servers and monitor replication/response; (2) stress test one server with up to 100 parallel clients over five minutes versus a Virtuoso SPARQL endpoint; (3) retrieve LIDDI 20 times normally and 20 times with a simulated unreliable connection (1% read errors), checking normalized equivalence, time, and failures.

Limitations

  • Stated: the current system and protocol are “not set in stone” and will need to evolve, especially network topology and partial replication, to handle thousands of servers and billions of nanopublications; the comparison to a SPARQL endpoint is “admittedly not a fair one”; the 1,000-reference index limit is “admittedly arbitrary”; replication restricts outgoing connections to one at a time, described as conservative.
  • Stated: “trustworthy” is used only in the narrow technical sense of verifying that an obtained file correctly represents a chosen version, not for judging whether a dataset is appropriate for a purpose.
  • Evident from the text: the first evaluation ran on only three servers, and the later network was 15 instances on 10 sites, so capacity estimates are extrapolations from a small deployment; no adversarial or security evaluation beyond hash verification is reported.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.7717/peerj-cs.78

Cites

40 of 40 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Verification

Claims the paper makes

0 of 13 cited sources checked · 0 supported · 0 flagged

  • Findingown data

    On average, 2,486 download attempts failed and had to be retried in the unreliable setting.

    “On average, 2,486 download attempts failed and had to be retried in the unreliable setting.”
    Quote in the paperNot a substantial claim
  • Findingreasoning

    The five nanopublications referred to by the index are already public.

    “There is no need to publish the five nanopublications this index is referring to, because they are already public (this is how we got them in the first place).”
    Quote in the paperSubstantial claim
  • Findingown data

    The SPARQL endpoint answering the same kind of requests needed 100 times longer to process them, 13 s on average.

    “In contrast, the SPARQL endpoint answering the same kind of requests needed 100 times longer to process them (13 s on average), consequently handled about 100 times fewer requests (1267), and started to hit the timeout of 60 s for some requests when more than 40 client accessed it in parallel.”
    Quote in the paperSubstantial claim
  • Findingown data

    The five nanopublication datasets used in the first part of the evaluation consist of a total of more than 5 million nanopublications and close to 200 million RDF triples, including nanopublication indexes generated for each dataset.

    “These five datasets consist of a total of more than 5 million nanopublications and close to 200 million RDF triples, including nanopublication indexes that we generated for each dataset.”
    Quote in the paperSubstantial claim
  • Findingown data

    The nanopublication server's average response time was 0.12 s.

    “The nanopublication server was able to handle 113,178 requests in total (i.e., an average of 377 requests per second) with an average response time of 0.12 s.”
    Quote in the paperSubstantial claim
  • It may state two separate claims as one. 36
  • Probably not a substantial claim. 36
  • The quote only partly states this claim. 1

  • It may state two separate claims as one.

    Findingown data

    The paper's evaluation of the current nanopublication server network shows that the system is efficient and reliable.

    “Our evaluation of the current network shows that this system is efficient and reliable.”
    Quote in the paperSubstantial claim

  • It may state two separate claims as one.

    Findingown data

    In the unreliable setting, the download times are more than doubled, with an average of 517 s and a maximum below 10 min.

    “the download times are more than doubled, but still in a very reasonable range with an average of 517 s and a maximum below 10 min.”
    Quote in the paperNot a substantial claim

  • It may state two separate claims as one.

    Findingown data

    In two of the 20 test runs, the connection was temporarily unreliable for natural reasons, and the number of download failures rose to 181 and 9,458 respectively.

    “In the remaining two runs, the connection happened to be temporarily unreliable for ''natural'' reasons, and the number of download failures rose to 181 and 9,458, respectively.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.

    Findingown data

    The first part of the evaluation lasted 13 h and 21 min, at which point all nanopublications were replicated on all three servers.

    “The first part of the evaluation lasted 13 h and 21 min, at which point all nanopublications were replicated on all three servers, and therefore the nanopublication traffic came to an end.”
    Quote in the paperNot a substantial claim

  • It may state two separate claims as one.

    Findingown data

    The nanopublication server handled 113,178 requests in total, an average of 377 requests per second.

    “The nanopublication server was able to handle 113,178 requests in total (i.e., an average of 377 requests per second) with an average response time of 0.12 s.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim. The paper’s conclusions may not depend on it.

    Findingown data

    Seven new sites in Amsterdam, Stanford, Barcelona, Ghent, Athens, Leipzig, and Haverford have joined the network since the first part of the evaluation.

    “Seven new sites in Amsterdam, Stanford, Barcelona, Ghent, Athens, Leipzig, and Haverford have joined the network since.”
    Quote in the paperNot a substantial claim

  • It may state two separate claims as one.

    Findingown data

    Not a single one of the 4,802 individual HTTP requests timed out, led to an error, or received a nanopublication that could not be successfully verified.

    “Not a single one of the 4,802 individual HTTP requests timed out, led to an error, or received a nanopublication that could not be successfully verified.”
    Quote in the paperNot a substantial claim

  • It may state two separate claims as one.

    Findingown data

    On average over the 20 test runs, the entire dataset was successfully downloaded in 235 s, with a maximum of 279 s.

    “On average over the 20 test runs, the entire dataset was successfully downloaded in 235 s, with a maximum of 279 s.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim. The paper’s conclusions may not depend on it.

    Findingown data

    Immediately after publication, only one server knows about the new nanopublication.

    “This is what you see immediately after publication.Only one server knows about the new nanopublication.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim. The paper’s conclusions may not depend on it.

    Findingown data

    Assigning trusty URIs to nanopublications does not change their content.

    “Looking into the file we can verify that nothing has changed with respect to the content”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim. The paper’s conclusions may not depend on it.

    Findingown data

    The current network consists of 15 server instances on 10 sites in 8 countries.

    “The current network of 15 server instances on 10 sites (in 8 countries) is shown in Fig. 5”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.

    Findingnone given

    After a few minutes, a published nanopublication will be distributed in the network and available on multiple servers.

    “Again, after just a few minutes this nanopublication will be distributed in the network and available on multiple servers.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim. The paper’s conclusions may not depend on it.

    Findingnone given

    At the time the first part of the evaluation was performed, the server network consisted of three servers in Zurich, New Haven, and Ottawa.

    “At the time the first part of the evaluation was performed, the server network consisted of three servers in Zurich, New Haven, and Ottawa.”
    Quote in the paperNot a substantial claim

  • It may state two separate claims as one.

    Findingown data

    The network handled an average of about 400,000 new nanopublications per hour (more than 100 per second), including loading each nanopublication once from the local file system, transferring it twice through the network, and verifying it three times.

    “The network was able to handle an average of about 400,000 new nanopublications per hour, which corresponds to more than 100 new nanopublications per second.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim. The paper’s conclusions may not depend on it.

    Findingown data

    The total size of the five datasets used in the first part of the evaluation, stored as uncompressed TriG files, is 15.6 GB.

    “The total size of these five datasets when stored as uncompressed TriG files amounts to 15.6 GB.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim.

    Premisecitation

    Traversal-based querying does not allow for the efficient and complete evaluation of certain types of queries.

    “Such limited interfaces normally have the drawback that traversal-based querying does not allow for the efficient and complete evaluation of certain types of queries (Hartig, 2013)”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Hartig, 2013 source

  • Probably not a substantial claim. The paper’s conclusions may not depend on it.

    Premisecitation

    Generating trusty URIs has minimal overhead for small files such as nanopublications, consisting only of about 1 millisecond per created nanopublication when the Java library is used.

    “For small files such as nanopublications, however, the overhead is minimal, consisting only of about 1 millisecond per created nanopublication when the Java library is used (Kuhn & Dumontier, 2014;Kuhn & Dumontier, 2015).”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Kuhn & Dumontier, 2014

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites Kuhn & Dumontier, 2015

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.

    Premisereasoning

    More light-weight interfaces such as read-write Linked Data Platform, Triple Pattern Fragments, and CumulusRDF constitute a scalability improvement but do not necessarily lead to perfect uptimes.

    “While this constitutes a scalability improvement (at the cost of, for instance, slower queries), it does not necessarily lead to perfect uptimes, as servers can be down for other reasons than excessive workload.”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites Speicher, Arwe & Malhotra, 2015

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites Verborgh et al., 2014 source
    • cites Ladwig & Harth, 2011

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • It may state two separate claims as one.

    Interpretationnone given

    Bundling all resources of scientific studies in large packages is not a necessity to ensure the availability of the involved resources and their robust interlinking; this can also be achieved with cryptographic identifiers and a decentralized architecture.

    “We demonstrate in this paper, however, that bundling all resources of scientific studies in large packages is not a necessity to ensure the availability of the involved resources and their robust interlinking, but we can achieve that also with cryptographic identifiers and a decentralized architecture.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim. The paper’s conclusions may not depend on it.

    Premisecitation

    Controlled natural language can serve in the fully formalized case as a user-friendly syntax for representing scientific facts.

    “we had shown before that controlled natural language can also serve in the fully formalized case as a user-friendly syntax for representing scientific facts (Kuhn et al., 2006).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Kuhn et al., 2006 source

  • Probably not a substantial claim. The paper may give no reason why it holds.

    Interpretationnone given

    In the proposed multi-layer architecture, the limitations of traversal-based querying do not apply because querying is only performed at a higher level where these limitations do not apply.

    “but this is not a problem with the multi-layer architecture we propose below, because querying is only performed at a higher level where these limitations do not apply.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim.

    Premisecitation

    The assertion of a nanopublication need not be fully formalized; informal or underspecified assertions can be allowed.

    “we argued in previous work that the assertion of a nanopublication need not be fully formalized, but we can allow for informal or underspecified assertions (Kuhn et al., 2013)”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Kuhn et al., 2013 source

  • Probably not a substantial claim. It may be too vague to check against evidence. The paper may give no reason why it holds.

    Premisenone given

    The creation of accurate semantic representations can be too challenging or too time-consuming for many scenarios and types of users, particularly in domains that lack ontologies and standardized terminologies with sufficient coverage.

    “to deal with the fact that the creation of accurate semantic representations can be too challenging or too time-consuming for many scenarios and types of users.This is particularly the case for domains that lack ontologies and standardized terminologies with sufficient coverage.”
    Quote in the paperNot a substantial claim

  • It may state two separate claims as one.

    Interpretationreasoning

    Because the hash is part of the URI, any link to such an artifact comes with the possibility to verify its content, including other trusty URI links it might contain, so the range of verifiability extends to the entire reference tree.

    “Because this hash is part of the URI, any link to such an artifact comes with the possibility to verify its content, including other trusty URI links it might contain.In this way, the range of verifiability extends to the entire reference tree.”
    Quote in the paperSubstantial claim

  • It may state two separate claims as one.

    Interpretationreasoning

    Together, the properties of nanopublications with trusty URIs significantly simplify the network's design and synchronization protocol and make it reliable and efficient even with limited resources.

    “Together, these aspects significantly simplify the design of such a network and its synchronization protocol, and make it reliable and efficient even with limited resources.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Nanopublications are always small by definition.
    • Rests on an assumption: Trusty URIs are guaranteed to be unique and universal.

  • Probably not a substantial claim.

    Interpretationreasoning

    The server components and their possible interactions allow for efficient decentralized distribution of published nanopublications.

    “Together, the server components and their possible interactions outlined above allow for efficient decentralized distribution of published nanopublications.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim. It may be too vague to check against evidence. It may state two separate claims as one.

    Interpretationreasoning

    We need to be able to group nanopublications and to identify and use large collections of them.

    “Therefore, we need to be able to group nanopublications and to identify and use large collections of them.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: Nanopublications are tiny atomic pieces of data.

  • Probably not a substantial claim. The paper may give no reason why it holds.

    Interpretationnone given

    Most processes will generate more than just one nanopublication, possibly thousands or even millions of them.

    “Similarly, most processes will generate more than just one nanopublication, possibly thousands or even millions of them.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: Nanopublications are tiny atomic pieces of data.

  • It may state two separate claims as one.

    Interpretationown data

    The load on a server, measured as response times, is barely noticeable for up to 50 parallel clients, and therefore the network can easily handle 50•x parallel client connections or more, where x is the number of independent physical servers in the network (currently x = 10).

    “The results of the second part show that the load on a server when measured as response times is barely noticeable for up to 50 parallel clients, and therefore the network can easily handle 50•x parallel client connections or more, where x is the number of independent physical servers in the network (currently x = 10).”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The per-server client capacity measured on the current network holds independently for each additional physical server, so that total capacity scales linearly with the number of servers.

  • Probably not a substantial claim. The paper may give no reason why it holds. It may state two separate claims as one.

    Interpretationnone given

    Nanopublications that have trusty URI identifiers can be uploaded to a server, distributed to other servers of the network, and retrieved from any of the servers or from multiple servers simultaneously, even if the original server is not accessible.

    “Nanopublications that have trusty URI identifiers can be uploaded to a server (or loaded from the local file system by the server administrator), and they are then distributed to the other servers of the network.They can then be retrieved from any of the servers, or from multiple servers simultaneously, even if the original server is not accessible.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim. It may be too vague to check against evidence. It may state two separate claims as one.

    Interpretationown data

    The presented server network can serve as a solid basis for semantic publishing, and possibly also for the Semantic Web in general.

    “We believe that the presented server network can serve as a solid basis for semantic publishing, and possibly also for the Semantic Web in general.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim. It may state two separate claims as one.

    Interpretationreasoning

    Analyses will mostly involve more than just one nanopublication and typically a large number of them.

    “This implies that analyses will mostly involve more than just one nanopublication and typically a large number of them.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Nanopublications are tiny atomic pieces of data.

  • It may state two separate claims as one.

    Interpretationreasoning

    If collection nanopublications contain other nanopublications, they would become very large for large collections and would quickly lose their property of being nano.

    “However, if we let such ''collection nanopublications'' contain other nanopublications, then the former would become very large for large collections and would quickly lose their property of being nano.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Nanopublications are tiny atomic pieces of data.

  • It may state two separate claims as one.

    Interpretationreasoning

    A trusty URI reference is much more reliable and more robust than links to other types of data repositories.

    “This reference is therefore much more reliable and more robust than links to other types of data repositories.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Links to other types of data repositories are less reliable and robust than trusty URI references.
    • Rests on an assumption: A URL will not resolve if the server it points to is temporarily or permanently down.

  • Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds. It may state two separate claims as one.

    Interpretationnone given

    The main limiting factor for scaling up the network is disk space, which is relatively cheap and easy to add.

    “The main limiting factor is disk space, which is relatively cheap and easy to add.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.

    Interpretationnone given

    Nanopublication indexes can have additional metadata attached, such as labels, descriptions, further references, and other types of relations at the level of an entire dataset.

    “In addition to describing sets of data entries, nanopublication indexes can also have additional metadata attached, such as labels, descriptions, further references, and other types of relations at the level of an entire dataset.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim. The paper’s conclusions may not depend on it.

    Interpretationreasoning

    A hash pattern like 'AA AB' would let a server replicate about 0.05% of all nanopublications.

    “As hashes are represented in Base64 notation, this particular hash pattern would let a server replicate about 0.05% of all nanopublications.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: Hashes are represented in Base64 notation.

  • It may state two separate claims as one.

    Premisenone given

    There is no standardized and generally accepted way of how to communicate mirror servers and how to decide on the client side whether a supposed mirror server is trustworthy.

    “there is no standardized and generally accepted way of how to communicate these mirror servers and how to decide on the client side whether a supposed mirror server is trustworthy.”
    Quote in the paperNot a substantial claim

  • It may state two separate claims as one.

    Interpretationreasoning

    A single point of failure can cause applications to be unable to complete their tasks: a single URI that does not resolve or a single server that does not respond can break the entire process.

    “A single point of failure can cause applications to be unable to complete their tasks: A single URI that does not resolve or a single server that does not respond can break the entire process.”
    Quote in the paperSubstantial claim

  • It may state two separate claims as one.

    Premisenone given

    All current solutions for Semantic Web applications are based on two-layer architectures and have no inherent replication mechanisms.

    “We can observe that all these current solutions are based on two-layer architectures, and have moreover no inherent replication mechanisms.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim. The paper may give no reason why it holds.

    Interpretationnone given

    The server network can be seen as an unstructured peer-to-peer network where each node freely decides which other nodes to connect to and which nanopublications to replicate.

    “The server network can be seen as an unstructured peer-to-peer network, where each node can freely decide which other nodes to connect to and which nanopublications to replicate.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim. It may state two separate claims as one.

    Interpretationnone given

    The nanopublication servers of the network connect to each other to retrieve and partly replicate their nanopublications, and they allow users to upload new nanopublications, which are then automatically distributed through the network.

    “The nanopublication servers of such a network connect to each other to retrieve and (partly) replicate their nanopublications, and they allow users to upload new nanopublications, which are then automatically distributed through the network.”
    Quote in the paperSubstantial claim

  • It may state two separate claims as one.

    Interpretationreasoning

    Some nanopublications may not be referenced by any index at all, while others may belong to several indexes at the same time.

    “We see that some nanopublications may not be referenced by any index at all, while others may belong to several indexes at the same time.”
    Quote in the paperNot a substantial claim

  • The quote only partly states this claim.

    Interpretationreasoning

    Basing the network's content on nanopublications with trusty URIs makes it easy to estimate the time needed to process an entity and the space needed to store it.

    “The first benefit is that the fact that nanopublications are always small (by definition) makes it easy to estimate how much time is needed to process an entity (such as validating its hash) and how much space to store it (e.g., as a serialized RDF string in a database).”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Nanopublications are always small by definition.

  • Probably not a substantial claim.

    Interpretationreasoning

    From this point on, everybody can conveniently and reliably retrieve the given set of nanopublications.

    “From this point on, everybody can conveniently and reliably retrieve the given set of nanopublications.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim. The paper may give no reason why it holds.

    Study design and conductnone given

    To ensure the immutability of nanopublications and to guarantee the reliability of the system, nanopublications are required to come with trusty URI identifiers.

    “To ensure the immutability of these nanopublications and to guarantee the reliability of the system, these nanopublications are required to come with trusty URI identifiers, i.e., they have to be transformed on the client side into such trusty nanopublications before they can be published to the network.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim. It may state two separate claims as one.

    Interpretationreasoning

    Nanopublication indexes have three types of relations: an index can append to another index, it can contain other indexes as sub-indexes, and it can contain nanopublications as elements.

    “In this way we end up with three types of relations: an index can append to another index, it can contain other indexes as sub-indexes, and it can contain nanopublications as elements.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.

    Study design and conductreasoning

    The comparison between the nanopublication server and the SPARQL endpoint is not a fair one, as SPARQL endpoints are much more powerful and are not tailor-made for the retrieval of nanopublications, though they provide a valuable and well-established reference point.

    “This comparison is admittedly not a fair one, as SPARQL endpoints are much more powerful and are not tailor-made for the retrieval of nanopublications, but they provide nevertheless a valuable and well-established reference point to evaluate the performance of our system.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim. It may be too vague to check against evidence. It may state two separate claims as one.

    Interpretationreasoning

    The most low-level task of providing Linked Data is essential for all other tasks at higher levels and therefore needs to be the most stable and robust one.

    “The most low-level task of providing Linked Data is essential for all other tasks at higher levels, and therefore needs to be the most stable and robust one.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim. The paper’s conclusions may not depend on it.

    Interpretationreasoning

    A set of 100,000 nanopublications can be defined by a sequence of 100 indexes, where each index appends to the previous and adds 1,000 nanopublications, together representing the entire set.

    “A set of 100,000 nanopublications, for example, can therefore be defined by a sequence of 100 indexes, where the first one stands for the first 1,000 nanopublications, the second one appends to the first and adds another 1,000 nanopublications (thereby representing 2,000 of them), and so on up to the last index, which appends to the second to last and thereby stands for the entire set.”
    Quote in the paperNot a substantial claim

  • It may state two separate claims as one.

    Interpretationnone given

    Approaches following the follow-your-nose principle can only be reliable and efficient if there is a guarantee that resolution of any single identifier will succeed within a short time frame and that processing the received representation will take up only a small amount of time and resources.

    “We argue that approaches following this principle can only be reliable and efficient if we have some sort of guarantee that the resolution of any single identifier will succeed within a short time frame in one way or another, and that the processing of the received representation will only take up a small amount of time and resources.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim. It may be too vague to check against evidence. The paper may give no reason why it holds. It may state two separate claims as one.

    Interpretationreasoning

    The presented approach could contribute to improve the availability and reproducibility of scientific results and put a reliable and trustworthy layer underneath the Semantic Web.

    “It could contribute to improve the availability and reproducibility of scientific results and put a reliable and trustworthy layer underneath the Semantic Web.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds. It may state two separate claims as one.

    Interpretationnone given

    The current implementation is designed to be run on normal web servers alongside other applications, with economic use of the server's resources in terms of memory and processing time.

    “The current implementation is designed to be run on normal web servers alongside with other applications, with economic use of the server's resources in terms of memory and processing time.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim. The paper’s conclusions may not depend on it.

    Interpretationnone given

    References of an index can also point to sub-indexes to allow datasets to be organized in hierarchies.

    “In addition, to allow datasets to be organized in hierarchies, we define that the references of an index can also point to sub-indexes.”
    Quote in the paperNot a substantial claim

  • It may state two separate claims as one.

    Interpretationown data

    The overall replication capacity of the current server network is around 9.4 million new nanopublications per day or 3.4 billion per year.

    “the first part of the evaluation shows that the overall replication capacity of the current server network is around 9.4 million new nanopublications per day or 3.4 billion per year.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The average replication rate observed during the 13-hour evaluation can be sustained over longer periods, so that it can be extrapolated to per-day and per-year capacity.

  • Probably not a substantial claim. It may be too vague to check against evidence. The paper may give no reason why it holds.

    Interpretationreasoning

    It is straightforward to represent collections of nanopublications as nanopublications themselves.

    “Given the versatility of the nanopublication standard, it seems straightforward to represent such collections as nanopublications themselves.”
    Quote in the paperNot a substantial claim

  • It may state two separate claims as one.

    Interpretationreasoning

    Nanopublications with trusty URIs are immutable and verifiable.

    “The third and possibly most important benefit is that nanopublications with trusty URIs are immutable and verifiable.”
    Quote in the paperSubstantial claim

  • It may state two separate claims as one.

    Interpretationreasoning

    Nanopublication servers can declare which subset of nanopublications they replicate and need to connect only to other servers whose subsets overlap.

    “Nanopublication servers are thereby given the opportunity to declare which subset of nanopublications they replicate, and need to connect only to those other servers whose subsets overlap.”
    Quote in the paperSubstantial claim

  • It may state two separate claims as one.

    Interpretationreasoning

    Because the protocol only allows for simple key-based lookup, the time complexity for all types of requests is sublinear and therefore scales up well.

    “As our protocol only allows for simple key-based lookup, the time complexity for all types of requests is sublinear and therefore scales up well.”
    Quote in the paperNot a substantial claim

  • It may state two separate claims as one.

    Interpretationreasoning

    To make follow-your-nose approaches reliable and efficient, RDF representations must be made available on several distributed servers and must be reasonably small.

    “This requires that (1) RDF representations are made available on several distributed servers, so the chance that they all happen to be inaccessible at the same time is negligible, and that (2) these representations are reasonably small, so that downloading them is a matter of fractions of a second, and so that one has to process only a reasonable amount of data to decide which links to follow.”
    Quote in the paperSubstantial claim
60 of 60

  • The paper relies on this without support.

    There exist currently no efficient, reliable, and agreed-upon methods for publishing scientific datasets.

    “there exist currently no efficient, reliable, and agreed-upon methods for publishing scientific datasets”

  • The paper relies on this without support.

    Modern science increasingly depends on datasets, which are left out in the classical way of publishing through narrative articles in journals or conference proceedings.

    “Modern science increasingly depends on datasets, which are however left out in the classical way of publishing, i.e., through narrative (printed or online) articles in journals or conference proceedings.”

  • The paper relies on this without support.

    The centralized nature of existing data repositories is inconsistent with the decentralized manner in which science is typically performed, and has serious consequences with respect to reliability and trust.

    “We argue that the centralized nature of existing data repositories is inconsistent with the decentralized manner in which science is typically performed, and that it has serious consequences with respect to reliability and trust.”

  • The paper relies on this without support.

    Datasets in RDF can be split into small pieces without any effects on their semantics; after Skolemization of blank nodes, RDF triples are independent and can be separated and joined without restrictions.

    “We furthermore exploit the fact that datasets in RDF can be split into small pieces without any effects on their semantics.After Skolemization of blank nodes, RDF triples are independent and can be separated and joined without restrictions.”

  • The paper relies on this without support.

    We can never perfectly trust any data we get.

    “We can therefore never perfectly trust any data we get, which significantly complicates the work of scientists and impedes the potential of fully automatic analyses.”

  • The paper relies on this without support.

    The creation of accurate semantic representations can be too challenging or too time-consuming for many scenarios and types of users, particularly in domains that lack ontologies and standardized terminologies with sufficient coverage.

    “the creation of accurate semantic representations can be too challenging or too time-consuming for many scenarios and types of users.This is particularly the case for domains that lack ontologies and standardized terminologies with sufficient coverage.”

  • The authors state this as an assumption.

    Nanopublications are always small by definition.

    “the fact that nanopublications are always small (by definition)”

  • The paper relies on this without support.

    Trusty URIs are guaranteed to be unique and universal.

    “the nanopublications already come with trusty URIs, which are guaranteed to be unique and universal”

  • The paper relies on this without support.

    The hashes of the current version of trusty URIs are 43 bytes long.

    “as the hashes of the current version of trusty URIs are 43 bytes long”

  • The paper relies on this without support.

    Hashes are represented in Base64 notation.

    “As hashes are represented in Base64 notation”

  • The paper relies on this without support.

    Nanopublications are tiny atomic pieces of data.

    “One of the core ideas behind nanopublications is that each of them is a tiny atomic piece of data.”

  • The paper relies on this without support.

    Trusty URIs allow reference links between nanopublications to be almost as strong as containment links.

    “trusty URIs allow us to make these reference links almost as strong as containment links”

  • The paper relies on this without support.

    Storage space is typically much easier and cheaper to scale up than memory or processing capacities.

    “storage space is typically much easier and cheaper to scale up than memory or processing capacities.”

  • The paper relies on this without support.

    The amount of nanopublications that can be expected to become publicly available in the next few years is large enough to be a meaningful load for the network.

    “To evaluate our approach, we want to find out whether a small server network run on normal web servers, without dedicated infrastructure, is able to handle the amount of nanopublications we can expect to become publicly available in the next few years.”

  • The paper relies on this without support.

    The datasets used for the first part of the evaluation are representative of the nanopublications that can be expected to become publicly available.

    “Table 1 Existing datasets in the nanopublication format, five of which were used for the first part of the evaluation.”

  • The paper relies on this without support.

    Links to other types of data repositories are less reliable and robust than trusty URI references.

    “This reference is therefore much more reliable and more robust than links to other types of data repositories.”

  • The paper relies on this without support.

    A URL will not resolve if the server it points to is temporarily or permanently down.

    “The URL will not resolve should the server be temporarily or permanently down”

  • The paper relies on this without support.

    The per-server client capacity measured on the current network holds independently for each additional physical server, so that total capacity scales linearly with the number of servers.

    “the network can easily handle 50•x parallel client connections or more, where x is the number of independent physical servers in the network (currently x = 10)”

  • The paper relies on this without support.

    The average replication rate observed during the 13-hour evaluation can be sustained over longer periods, so that it can be extrapolated to per-day and per-year capacity.

    “the overall replication capacity of the current server network is around 9.4 million new nanopublications per day or 3.4 billion per year”
19 of 19

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • LIDDI triple count conflicts with Table 1

    Minor

    The evaluation design says LIDDI has around 100,000 triples, but Table 1 lists 2,051,959 content triples and 99,272 index triples. Around 100,000 describes its nanopublication count, not its triple count.

1 of 1

Citations

Cites

Flagged first, then by citations

2 of the 42 references couldn’t be matched to a known paper.

Cited by

Sorted by citations

Related

Metadata

Authors
Kuhn, T., Chichester, C., Krauthammer, M., Queralt-Rosinach, N., Verborgh, R., Giannakopoulos, G., Ngonga Ngomo, A.-C., Viglianti, R., & Dumontier, M.
Affiliations
  • Vrije Universiteit Amsterdam (NL)
  • Nestle Institute of Health Sciences, Lausanne, Switzerland
  • Yale University (US)
  • Research Programme on Biomedical Informatics, Hospital del Mar Medical Research Institute, Universitat Pompeu Fabra, Barcelona, Spain
  • Ghent University (BE)
  • National Centre of Scientific Research "Demokritos" (GR)
  • AKSW Research Group, University of Leipzig, Leipzig, Germany
  • University of Maryland, College Park (US)
  • Stanford University (US)
ORCID
Venue
PeerJ Computer Science
Publisher
PeerJ
Published
2016-08-22
Volume
2
Pages
e78
ISSN
2376-5992 (ISSN-L)
DOI
10.7717/peerj-cs.78
OpenAlex
W2284922289
Cited by
90
References
42
Citations by year
2026: 7 · 2025: 4 · 2024: 7 · 2023: 5 · 2022: 7 · 2021: 21 · 2020: 10 · 2019: 7 · 2018: 9 · 2017: 11 · 2016: 2
FWCI
7.14
Type
journal article
Language
en
Open access
Licence
Locations
  • PeerJ Computer Science (publishedVersion, cc-by)
  • Ghent University Academic Bibliography (Ghent University) (submittedVersion, cc-by)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
  • Repositori digital de la UPF (Universitat Pompeu Fabra) (submittedVersion, cc-by)
  • VU Research Portal (submittedVersion, other-oa)
  • Data Archiving and Networked Services (DANS) (submittedVersion, other-oa)
Funders
Research Foundation–Flanders (FWO)
Topics
  • Physical Sciences › Computer Science › Artificial Intelligence › Semantic Web and Ontologies
  • Physical Sciences › Computer Science › Information Systems › Research Data Management Practices
  • Social Sciences › Decision Sciences › Information Systems and Management › Scientific Computing and Data Management
Keywords
nanopublications, data publication, decentralized storage, provenance, RDF, semantic web, data archiving, server networks, data verification

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

A language model drew the claims from the paper on 2026-11-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-11-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review