Linkitup: Link Discovery for Research Data
6 likes•3,855 views
The document discusses the critical importance of data sharing and preservation in scientific research, emphasizing that research success increasingly depends on accessible data. It highlights the existing shortcomings in data management across various disciplines and calls for immediate action from funding agencies, universities, and researchers to establish robust frameworks for open data access and support. The need for metadata management software and educational programs on data management is also stressed to ensure effective reuse and understanding of data throughout its life cycle.
![Data is the Bottleneck
Common Motifs in Scientific Workflows:
An Empirical Analysis
Daniel Garijo⇤ , Pinar Alper † , Khalid Belhajjame† , Oscar Corcho⇤ , Yolanda Gil‡ , Carole Goble†
⇤ Ontology
Engineering Group, Universidad Polit´ cnica de Madrid. {dgarijo, ocorcho}@fi.upm.es
e
of Computer Science, University of Manchester. {alperp, khalidb, carole.goble}@cs.manchester.ac.uk
‡ Information Sciences Institute, Department of Computer Science, University of Southern California. gil@isi.edu
† School
Abstract—While workflow technology has gained momentum
in the last decade as a means for specifying and enacting computational experiments in modern science, reusing and repurposing
existing workflows to build new scientific experiments is still a
daunting task. This is partly due to the difficulty that scientists
experience when attempting to understand existing workflows,
which contain several data preparation and adaptation steps in
addition to the scientifically significant analysis steps. One way
to tackle the understandability problem is through providing
abstractions that give a high-level view of activities undertaken
within workflows. As a first step towards abstractions, we report
in this paper on the results of a manual analysis performed over
a set of real-world scientific workflows from Taverna and Wings
systems. Our analysis has resulted in a set of scientific workflow
motifs that outline i) the kinds of data intensive activities that are
observed in workflows (data oriented motifs), and ii) the different
manners in which activities are implemented within workflows
(workflow oriented motifs). These motifs can be useful to inform
workflow designers on the good and bad practices for workflow
development, to inform the design of automated tools for the
generation of workflow abstractions, etc.
I. I NTRODUCTION
Scientific workflows have been increasingly used in the last
decade as an instrument for data intensive scientific analysis.
In these settings, workflows serve a dual function: first as
detailed documentation of the method (i. e. the input sources
and processing steps taken for the derivation of a certain
data item) and second as re-usable, executable artifacts for
data-intensive analysis. Workflows stitch together a variety
of data manipulation activities such as data movement, data
transformation or data visualization to serve the goals of the
scientific study. The stitching is realized by the constructs
made available by the workflow system used and is largely
shaped by the environment in which the system operates and
the function undertaken by the workflow.
A variety of workflow systems are in use [10] [3] [7] [2]
serving several scientific disciplines. A workflow is a software
artifact, and as such once developed and tested, it can be
shared and exchanged between scientists. Other scientists can
then reuse existing workflows in their experiments, e.g., as
sub-workflows [17]. Workflow reuse presents several advantages [4]. For example, it enables proper data citation and
improves quality through shared workflow development by
leveraging the expertise of previous users. Users can also
re-purpose existing workflows to adapt them to their needs
[4]. Emerging workflow repositories such as myExperiment
[14] and CrowdLabs [8] have made publishing and finding
workflows easier, but scientists still face the challenges of reuse, which amounts to fully understanding and exploiting the
available workflows/fragments. One difficulty in understanding
workflows is their complex nature. A workflow may contain
several scientifically-significant analysis steps, combined with
various other data preparation activities, and in different
implementation styles depending on the environment and
context in which the workflow is executed. The difficulty in
understanding causes workflow developers to revert to starting
from scratch rather than re-using existing fragments.
Through an analysis of the current practices in scientific
workflow development, we could gain insights on the creation
of understandable and more effectively re-usable workflows.
Specifically, we propose an analysis with the following objectives:
1) To reverse-engineer the set of current practices in workflow development through an analysis of empirical evidence.
2) To identify workflow abstractions that would facilitate
understandability and therefore effective re-use.
3) To detect potential information sources and heuristics
that can be used to inform the development of tools for
creating workflow abstractions.
In this paper we present the result of an empirical analysis
performed over 177 workflow descriptions from Taverna [10]
and Wings [3]. Based on this analysis, we propose a catalogue
of scientific workflow motifs. Motifs are provided through i)
a characterization of the kinds of data-oriented activities that
are carried out within workflows, which we refer to as dataoriented motifs, and ii) a characterization of the different manners in which those activity motifs are realized/implemented
within workflows, which we refer to as workflow-oriented
motifs. It is worth mentioning that, although important, motifs
that have to do with scheduling and mapping of workflows
onto distributed resources [12] are out the scope of this paper.
The paper is structured as follows. We begin by providing
related work in Section II, which is followed in Section III by
brief background information on Scientific Workflows, and the
two systems that were subject to our analysis. Afterwards we
describe the dataset and the general approach of our analysis.
We present the detected scientific workflow motifs in Section
IV and we highlight the main features of their distribution](https://image.slidesharecdn.com/hoekstra-groth-linkitup-aaai2013-fall-131116102511-phpapp01/85/Linkitup-Link-Discovery-for-Research-Data-13-320.jpg)
![Data is the Bottleneck
Common Motifs in Scientific Workflows:
An Empirical Analysis
Daniel Garijo⇤ , Pinar Alper † , Khalid Belhajjame† , Oscar Corcho⇤ , Yolanda Gil‡ , Carole Goble†
⇤ Ontology
Engineering Group, Universidad Polit´ cnica de Madrid. {dgarijo, ocorcho}@fi.upm.es
e
of Computer Science, University of Manchester. {alperp, khalidb, carole.goble}@cs.manchester.ac.uk
‡ Information Sciences Institute, Department of Computer Science, University of Southern California. gil@isi.edu
† School
Abstract—While workflow technology has gained momentum
in the last decade as a means for specifying and enacting computational experiments in modern science, reusing and repurposing
existing workflows to build new scientific experiments is still a
daunting task. This is partly due to the difficulty that scientists
experience when attempting to understand existing workflows,
which contain several data preparation and adaptation steps in
addition to the scientifically significant analysis steps. One way
to tackle the understandability problem is through providing
abstractions that give a high-level view of activities undertaken
within workflows. As a first step towards abstractions, we report
in this paper on the results of a manual analysis performed over
a set of real-world scientific workflows from Taverna and Wings
systems. Our analysis has resulted in a set of scientific workflow
motifs that outline i) the kinds of data intensive activities that are
observed in workflows (data oriented motifs), and ii) the different
manners in which activities are implemented within workflows
(workflow oriented motifs). These motifs can be useful to inform
workflow designers on the good and bad practices for workflow
development, to inform the design of automated tools for the
generation of workflow abstractions, etc.
I. I NTRODUCTION
Scientific workflows have been increasingly used in the last
decade as an instrument for data intensive scientific analysis.
In these settings, workflows serve a dual function: first as
detailed documentation of the method (i. e. the input sources
and processing steps taken for the derivation of a certain
data item) and second as re-usable, executable artifacts for
data-intensive analysis. Workflows stitch together a variety
of data manipulation activities such as data movement, data
transformation or data visualization to serve the goals of the
scientific study. The stitching is realized by the constructs
made available by the workflow system used and is largely
shaped by the environment in which the system operates and
the function undertaken by the workflow.
A variety of workflow systems are in use [10] [3] [7] [2]
serving several scientific disciplines. A workflow is a software
artifact, and as such once developed and tested, it can be
shared and exchanged between scientists. Other scientists can
then reuse existing workflows in their experiments, e.g., as
sub-workflows [17]. Workflow reuse presents several advantages [4]. For example, it enables proper data citation and
improves quality through shared workflow development by
leveraging the expertise of previous users. Users can also
re-purpose existing workflows to adapt them to their needs
[4]. Emerging workflow repositories such as myExperiment
[14] and CrowdLabs [8] have made publishing and finding
workflows easier, but scientists still face the challenges of reuse, which amounts to fully understanding and exploiting the
available workflows/fragments. One difficulty in understanding
workflows is their complex nature. A workflow may contain
several scientifically-significant analysis steps, combined with
various other data preparation activities, and in different
implementation styles depending on the environment and
context in which the workflow is executed. The difficulty in
understanding causes workflow developers to revert to starting
from scratch rather than re-using existing fragments.
Through an analysis of the current practices in scientific
workflow development, we could gain insights on the creation
of understandable and more effectively re-usable workflows.
Specifically, we propose an analysis with the following objectives:
1) To reverse-engineer the set of current practices in workflow development through an analysis of empirical evidence.
2) To identify workflow abstractions that would facilitate
understandability and therefore effective re-use.
3) To detect potential information sources and heuristics
that can be used to inform the development of tools for
creating workflow abstractions.
In this paper we present the result of an empirical analysis
performed over 177 workflow descriptions from Taverna [10]
and Wings [3]. Based on this analysis, we propose a catalogue
of scientific workflow motifs. Motifs are provided through i)
a characterization of the kinds of data-oriented activities that
are carried out within workflows, which we refer to as dataoriented motifs, and ii) a characterization of the different manners in which those activity motifs are realized/implemented
within workflows, which we refer to as workflow-oriented
motifs. It is worth mentioning that, although important, motifs
that have to do with scheduling and mapping of workflows
onto distributed resources [12] are out the scope of this paper.
The paper is structured as follows. We begin by providing
related work in Section II, which is followed in Section III by
brief background information on Scientific Workflows, and the
two systems that were subject to our analysis. Afterwards we
describe the dataset and the general approach of our analysis.
We present the detected scientific workflow motifs in Section
IV and we highlight the main features of their distribution
Data-Oriented Motifs per Domain
Fig. 3.
Distribution of Data-Oriented Motifs per domain](https://image.slidesharecdn.com/hoekstra-groth-linkitup-aaai2013-fall-131116102511-phpapp01/85/Linkitup-Link-Discovery-for-Research-Data-14-320.jpg)
![Data is the Bottleneck
Common Motifs in Scientific Workflows:
An Empirical Analysis
Daniel Garijo⇤ , Pinar Alper † , Khalid Belhajjame† , Oscar Corcho⇤ , Yolanda Gil‡ , Carole Goble†
⇤ Ontology
Engineering Group, Universidad Polit´ cnica de Madrid. {dgarijo, ocorcho}@fi.upm.es
e
of Computer Science, University of Manchester. {alperp, khalidb, carole.goble}@cs.manchester.ac.uk
‡ Information Sciences Institute, Department of Computer Science, University of Southern California. gil@isi.edu
† School
Abstract—While workflow technology has gained momentum
in the last decade as a means for specifying and enacting computational experiments in modern science, reusing and repurposing
existing workflows to build new scientific experiments is still a
daunting task. This is partly due to the difficulty that scientists
experience when attempting to understand existing workflows,
which contain several data preparation and adaptation steps in
addition to the scientifically significant analysis steps. One way
to tackle the understandability problem is through providing
abstractions that give a high-level view of activities undertaken
within workflows. As a first step towards abstractions, we report
in this paper on the results of a manual analysis performed over
a set of real-world scientific workflows from Taverna and Wings
systems. Our analysis has resulted in a set of scientific workflow
motifs that outline i) the kinds of data intensive activities that are
observed in workflows (data oriented motifs), and ii) the different
manners in which activities are implemented within workflows
(workflow oriented motifs). These motifs can be useful to inform
workflow designers on the good and bad practices for workflow
development, to inform the design of automated tools for the
generation of workflow abstractions, etc.
Fig. 3.
[14] and CrowdLabs [8] have made publishing and finding
workflows easier, but scientists still face the challenges of reuse, which amounts to fully understanding and exploiting the
available workflows/fragments. One difficulty in understanding
workflows is their complex nature. A workflow may contain
several scientifically-significant analysis steps, combined with
various other data preparation activities, and in different
implementation styles depending on the environment and
context in which the workflow is executed. The difficulty in
understanding causes workflow developers to revert to starting
from scratch rather than re-using existing fragments.
Through an analysis of the current practices in scientific
workflow development, we could gain insights on the creation
of understandable and more effectively re-usable workflows.
Specifically, we propose an analysis with the following objectives:
Distribution of Data-Orientedpractices in work- domain
1) To reverse-engineer the set of current Motifs per
I. I NTRODUCTION
Scientific workflows have been increasingly used in the last
decade as an instrument for data intensive scientific analysis.
In these settings, workflows serve a dual function: first as
detailed documentation of the method (i. e. the input sources
and processing steps taken for the derivation of a certain
data item) and second as re-usable, executable artifacts for
data-intensive analysis. Workflows stitch together a variety
of data manipulation activities such as data movement, data
transformation or data visualization to serve the goals of the
scientific study. The stitching is realized by the constructs
made available by the workflow system used and is largely
shaped by the environment in which the system operates and
the function undertaken by the workflow.
A variety of workflow systems are in use [10] [3] [7] [2]
serving several scientific disciplines. A workflow is a software
artifact, and as such once developed and tested, it can be
shared and exchanged between scientists. Other scientists can
then reuse existing workflows in their experiments, e.g., as
sub-workflows [17]. Workflow reuse presents several advantages [4]. For example, it enables proper data citation and
improves quality through shared workflow development by
leveraging the expertise of previous users. Users can also
re-purpose existing workflows to adapt them to their needs
[4]. Emerging workflow repositories such as myExperiment
flow development through an analysis of empirical evidence.
2) To identify workflow abstractions that would facilitate
understandability and therefore effective re-use.
3) To detect potential information sources and heuristics
that can be used to inform the development of tools for
creating workflow abstractions.
In this paper we present the result of an empirical analysis
performed over 177 workflow descriptions from Taverna [10]
and Wings [3]. Based on this analysis, we propose a catalogue
of scientific workflow motifs. Motifs are provided through i)
a characterization of the kinds of data-oriented activities that
are carried out within workflows, which we refer to as dataoriented motifs, and ii) a characterization of the different manners in which those activity motifs are realized/implemented
within workflows, which we refer to as workflow-oriented
motifs. It is worth mentioning that, although important, motifs
that have to do with scheduling and mapping of workflows
onto distributed resources [12] are out the scope of this paper.
The paper is structured as follows. We begin by providing
related work in Section II, which is followed in Section III by
brief background information on Scientific Workflows, and the
two systems that were subject to our analysis. Afterwards we
describe the dataset and the general approach of our analysis.
We present the detected scientific workflow motifs in Section
IV and we highlight the main features of their distribution
Fig. 5.
Fig. 3.
Data-Preparation Motifs per Domain
Data-Oriented Motifs per Domain
Data Preparation Motifs in the Genomics Wo
Distribution of Data-Oriented Motifs per domain](https://image.slidesharecdn.com/hoekstra-groth-linkitup-aaai2013-fall-131116102511-phpapp01/85/Linkitup-Link-Discovery-for-Research-Data-15-320.jpg)
Linkitup: Link Discovery for Research Data
- 1. 2 Semantics Datato From Data Semantics for Scientific Data Publishers linkitup Link Discovery for Research Data Rinke Hoekstra and Paul Groth Network Insitute, VU University Amsterdam Law Faculty, University of Amsterdam ★ ★ Linkitup - Link Discovery for Research Data by Rinke Hoekstra Licensed under a Creative Commons Attribution-ShareAlike 3.0 Unported License.
- 2. 2 Semantics Datato From Data Semantics for Scientific Data Publishers linkitup Link Discovery for Research Data Rinke Hoekstra and Paul Groth Network Insitute, VU University Amsterdam Law Faculty, University of Amsterdam ★ ★ How to share, publish, access, analyse, interpret and reuse data? Linkitup - Link Discovery for Research Data by Rinke Hoekstra Licensed under a Creative Commons Attribution-ShareAlike 3.0 Unported License.
- 8. www.nature.com/nature Data’s shameful neglect Vol 461 | Issue no. 7261 | 10 September 2009 Research cannot flourish if data are not preserved and made accessible. All concerned must act accordingly. M ore and more often these days, a research project’s success is measured not just by the publications it produces, but also by the data it makes available to the wider community. Pioneering archives such as GenBank have demonstrated just how powerful such legacy data sets can be for generating new discoveries — especially when data are combined from many laboratories and analysed in ways that the original researchers could not have anticipated. All but a handful of disciplines still lack the technical, institutional and cultural frameworks required to support such open data access (see pages 168 and 171) — leading to a scandalous shortfall in the sharing of data by researchers (see page 160). This deficiency urgently needs to be addressed by funders, universities and the researchers themselves. Research funding agencies need to recognize that preservation of and access to digital data are central to their mission, and need to be supported accordingly. Organizations in the United Kingdom, for instance, have made a good start. The Joint Information Systems Committee, established by the seven UK research councils in 1993, has made data-sharing a priority, and has helped to establish a Digital Curation Centre, headquartered at the University of Edinburgh, to be a national focus for research and development into data issues. Other European agencies have also pursued initiatives. The United States, by contrast, is playing catch-up. Since 2005, a 29-member Interagency Working Group on Digital Data has been trying to get US funding agencies to develop plans for how they will support data archiving — and just as importantly, to develop policies on what data should and should not be preserved, and what exceptions should be made for reasons such as patient privacy. Some agencies have taken the lead in doing so; many more are hanging back. They should all being moving forwards vigorously. What is more, funding agencies and researchers alike must ensure that they support not only the hardware needed to store the data, but also the software that will help investigators to do this. One important facet is metadata management software: tools that streamline the tedious process of annotating data with a description of what the bits mean, which instrument collected them, which algorithms have been used to process them and so on — information that is essential if other scientists are to reuse the data effectively. Also necessary, especially in an era when data can be mixed and combined in unanticipated ways, is software that can keep track of which pieces of data came from whom. Such systems are essential if tenure and promotion committees are ever to give credit — as they should — to candidates’ track-record of “Data management data contribution. Who should host these data? Agencies should be woven and the research community together into every course in need to create the digital equivalent science.” of libraries: institutions that can take responsibility for preserving digital data and making them accessible over the long term. The university research libraries themselves are obvious candidates to assume this role. But whoever takes it on, data preservation will require robust, long-term funding. One potentially helpful initiative is the US National Science Foundation’s DataNet programme, in which researchers are exploring financial mechanisms such as subscription services and membership fees. Finally, universities and individual disciplines need to undertake a vigorous programme of education and outreach about data. Consider, for example, that most university science students get a reasonably good grounding in statistics. But their studies rarely include anything about information management — a discipline that encompasses the entire life cycle of data, from how they are acquired and stored to how they are organized, retrieved and maintained over time. That needs to change: data management should be woven into every course in science, as one of the foundations of knowledge. ■ A step too far? a base on the Moon, then send them to Mars. This idea immediately set off a debate that is still continuing, in which sceptics ask whether there is any point in returning to the Moon nearly half a century after the first landings. Why not go to Mars directly, or visit nearEarth asteroids, or send people to service telescopes in the deep space beyond Earth? Yet that debate is both counter-productive — a new set of rockets could go to all of these places — and moot, because Bush’s vision never attracted the hoped-for budget increases. Indeed, a blue-riband commission reporting to US President Barack Obama this week (see page 153) finds the organizational malaise unchanged: NASA is still doing too much with too little. Without more money, the agency won’t be sending people anywhere beyond the International Space Station, which resides in low Earth orbit only 350 kilometres up. And even the ability to do that is in question: Ares I, the US rocket that would return Research cannot flourish if data are not preserved and made accessible. All concerned must act accordingly. DATA The Obama administration must fund human space flight adequately, or stop speaking of ‘exploration’. A fter the space shuttle Columbia burned up during re-entry into Earth’s atmosphere in 2003, the board that was convened to investigate the disaster looked beyond its technical causes to NASA’s organizational malaise. For decades, the board pointed out, the shuttle programme had been trying to do too much with too little money. NASA desperately needed a clearer vision and a better-defined mission for human space flight. The next year, then-President George W. Bush attempted to supply that vision with a new long-term goal: first send astronauts to build 145 145-146 Editorials WF IF.indd 145 8/9/09 14:06:40 Silver Bullet? http://on.wsj.com/XCajtB
- 9. www.nature.com/nature Data’s shameful neglect Vol 461 | Issue no. 7261 | 10 September 2009 Research cannot flourish if data are not preserved and made accessible. All concerned must act accordingly. M ore and more often these days, a research project’s success is measured not just by the publications it produces, but also by the data it makes available to the wider community. Pioneering archives such as GenBank have demonstrated just how powerful such legacy data sets can be for generating new discoveries — especially when data are combined from many laboratories and analysed in ways that the original researchers could not have anticipated. All but a handful of disciplines still lack the technical, institutional and cultural frameworks required to support such open data access (see pages 168 and 171) — leading to a scandalous shortfall in the sharing of data by researchers (see page 160). This deficiency urgently needs to be addressed by funders, universities and the researchers themselves. Research funding agencies need to recognize that preservation of and access to digital data are central to their mission, and need to be supported accordingly. Organizations in the United Kingdom, for instance, have made a good start. The Joint Information Systems Committee, established by the seven UK research councils in 1993, has made data-sharing a priority, and has helped to establish a Digital Curation Centre, headquartered at the University of Edinburgh, to be a national focus for research and development into data issues. Other European agencies have also pursued initiatives. The United States, by contrast, is playing catch-up. Since 2005, a 29-member Interagency Working Group on Digital Data has been trying to get US funding agencies to develop plans for how they will support data archiving — and just as importantly, to develop policies on what data should and should not be preserved, and what exceptions should be made for reasons such as patient privacy. Some agencies have taken the lead in doing so; many more are hanging back. They should all being moving forwards vigorously. What is more, funding agencies and researchers alike must ensure that they support not only the hardware needed to store the data, but also the software that will help investigators to do this. One important facet is metadata management software: tools that streamline the tedious process of annotating data with a description of what the bits mean, which instrument collected them, which algorithms have been used to process them and so on — information that is essential if other scientists are to reuse the data effectively. Also necessary, especially in an era when data can be mixed and combined in unanticipated ways, is software that can keep track of which pieces of data came from whom. Such systems are essential if tenure and promotion committees are ever to give credit — as they should — to candidates’ track-record of “Data management data contribution. Who should host these data? Agencies should be woven and the research community together into every course in need to create the digital equivalent science.” of libraries: institutions that can take responsibility for preserving digital data and making them accessible over the long term. The university research libraries themselves are obvious candidates to assume this role. But whoever takes it on, data preservation will require robust, long-term funding. One potentially helpful initiative is the US National Science Foundation’s DataNet programme, in which researchers are exploring financial mechanisms such as subscription services and membership fees. Finally, universities and individual disciplines need to undertake a vigorous programme of education and outreach about data. Consider, for example, that most university science students get a reasonably good grounding in statistics. But their studies rarely include anything about information management — a discipline that encompasses the entire life cycle of data, from how they are acquired and stored to how they are organized, retrieved and maintained over time. That needs to change: data management should be woven into every course in science, as one of the foundations of knowledge. ■ A step too far? a base on the Moon, then send them to Mars. This idea immediately set off a debate that is still continuing, in which sceptics ask whether there is any point in returning to the Moon nearly half a century after the first landings. Why not go to Mars directly, or visit nearEarth asteroids, or send people to service telescopes in the deep space beyond Earth? Yet that debate is both counter-productive — a new set of rockets could go to all of these places — and moot, because Bush’s vision never attracted the hoped-for budget increases. Indeed, a blue-riband commission reporting to US President Barack Obama this week (see page 153) finds the organizational malaise unchanged: NASA is still doing too much with too little. Without more money, the agency won’t be sending people anywhere beyond the International Space Station, which resides in low Earth orbit only 350 kilometres up. And even the ability to do that is in question: Ares I, the US rocket that would return Research cannot flourish if data are not preserved and made accessible. All concerned must act accordingly. DATA The Obama administration must fund human space flight adequately, or stop speaking of ‘exploration’. A fter the space shuttle Columbia burned up during re-entry into Earth’s atmosphere in 2003, the board that was convened to investigate the disaster looked beyond its technical causes to NASA’s organizational malaise. For decades, the board pointed out, the shuttle programme had been trying to do too much with too little money. NASA desperately needed a clearer vision and a better-defined mission for human space flight. The next year, then-President George W. Bush attempted to supply that vision with a new long-term goal: first send astronauts to build 145 145-146 Editorials WF IF.indd 145 8/9/09 14:06:40 Silver Bullet? http://on.wsj.com/XCajtB
- 10. www.nature.com/nature Data’s shameful neglect Vol 461 | Issue no. 7261 | 10 September 2009 Research cannot flourish if data are not preserved and made accessible. All concerned must act accordingly. M ore and more often these days, a research project’s success is measured not just by the publications it produces, but also by the data it makes available to the wider community. Pioneering archives such as GenBank have demonstrated just how powerful such legacy data sets can be for generating new discoveries — especially when data are combined from many laboratories and analysed in ways that the original researchers could not have anticipated. All but a handful of disciplines still lack the technical, institutional and cultural frameworks required to support such open data access (see pages 168 and 171) — leading to a scandalous shortfall in the sharing of data by researchers (see page 160). This deficiency urgently needs to be addressed by funders, universities and the researchers themselves. Research funding agencies need to recognize that preservation of and access to digital data are central to their mission, and need to be supported accordingly. Organizations in the United Kingdom, for instance, have made a good start. The Joint Information Systems Committee, established by the seven UK research councils in 1993, has made data-sharing a priority, and has helped to establish a Digital Curation Centre, headquartered at the University of Edinburgh, to be a national focus for research and development into data issues. Other European agencies have also pursued initiatives. The United States, by contrast, is playing catch-up. Since 2005, a 29-member Interagency Working Group on Digital Data has been trying to get US funding agencies to develop plans for how they will support data archiving — and just as importantly, to develop policies on what data should and should not be preserved, and what exceptions should be made for reasons such as patient privacy. Some agencies have taken the lead in doing so; many more are hanging back. They should all being moving forwards vigorously. What is more, funding agencies and researchers alike must ensure that they support not only the hardware needed to store the data, but also the software that will help investigators to do this. One important facet is metadata management software: tools that streamline the tedious process of annotating data with a description of what the bits mean, which instrument collected them, which algorithms have been used to process them and so on — information that is essential if other scientists are to reuse the data effectively. Also necessary, especially in an era when data can be mixed and combined in unanticipated ways, is software that can keep track of which pieces of data came from whom. Such systems are essential if tenure and promotion committees are ever to give credit — as they should — to candidates’ track-record of “Data management data contribution. Who should host these data? Agencies should be woven and the research community together into every course in need to create the digital equivalent science.” of libraries: institutions that can take responsibility for preserving digital data and making them accessible over the long term. The university research libraries themselves are obvious candidates to assume this role. But whoever takes it on, data preservation will require robust, long-term funding. One potentially helpful initiative is the US National Science Foundation’s DataNet programme, in which researchers are exploring financial mechanisms such as subscription services and membership fees. Finally, universities and individual disciplines need to undertake a vigorous programme of education and outreach about data. Consider, for example, that most university science students get a reasonably good grounding in statistics. But their studies rarely include anything about information management — a discipline that encompasses the entire life cycle of data, from how they are acquired and stored to how they are organized, retrieved and maintained over time. That needs to change: data management should be woven into every course in science, as one of the foundations of knowledge. ■ A step too far? a base on the Moon, then send them to Mars. This idea immediately set off a debate that is still continuing, in which sceptics ask whether there is any point in returning to the Moon nearly half a century after the first landings. Why not go to Mars directly, or visit nearEarth asteroids, or send people to service telescopes in the deep space beyond Earth? Yet that debate is both counter-productive — a new set of rockets could go to all of these places — and moot, because Bush’s vision never attracted the hoped-for budget increases. Indeed, a blue-riband commission reporting to US President Barack Obama this week (see page 153) finds the organizational malaise unchanged: NASA is still doing too much with too little. Without more money, the agency won’t be sending people anywhere beyond the International Space Station, which resides in low Earth orbit only 350 kilometres up. And even the ability to do that is in question: Ares I, the US rocket that would return Research cannot flourish if data are not preserved and made accessible. All concerned must act accordingly. DATA The Obama administration must fund human space flight adequately, or stop speaking of ‘exploration’. A fter the space shuttle Columbia burned up during re-entry into Earth’s atmosphere in 2003, the board that was convened to investigate the disaster looked beyond its technical causes to NASA’s organizational malaise. For decades, the board pointed out, the shuttle programme had been trying to do too much with too little money. NASA desperately needed a clearer vision and a better-defined mission for human space flight. The next year, then-President George W. Bush attempted to supply that vision with a new long-term goal: first send astronauts to build 145 145-146 Editorials WF IF.indd 145 8/9/09 14:06:40 Silver Bullet? http://on.wsj.com/XCajtB
- 11. Repository Services • • • • • Data is easy to upload Landing page for data Citable reference for data Default licensing options Guarantees for long term archival
- 12. Standard Metadata • Provenance metadata • Content metadata • • Metadata is locked in authors, title, publication date free text tags, categories, links Hard to interpret the data itself
- 13. Data is the Bottleneck Common Motifs in Scientific Workflows: An Empirical Analysis Daniel Garijo⇤ , Pinar Alper † , Khalid Belhajjame† , Oscar Corcho⇤ , Yolanda Gil‡ , Carole Goble† ⇤ Ontology Engineering Group, Universidad Polit´ cnica de Madrid. {dgarijo, ocorcho}@fi.upm.es e of Computer Science, University of Manchester. {alperp, khalidb, carole.goble}@cs.manchester.ac.uk ‡ Information Sciences Institute, Department of Computer Science, University of Southern California. [email protected] † School Abstract—While workflow technology has gained momentum in the last decade as a means for specifying and enacting computational experiments in modern science, reusing and repurposing existing workflows to build new scientific experiments is still a daunting task. This is partly due to the difficulty that scientists experience when attempting to understand existing workflows, which contain several data preparation and adaptation steps in addition to the scientifically significant analysis steps. One way to tackle the understandability problem is through providing abstractions that give a high-level view of activities undertaken within workflows. As a first step towards abstractions, we report in this paper on the results of a manual analysis performed over a set of real-world scientific workflows from Taverna and Wings systems. Our analysis has resulted in a set of scientific workflow motifs that outline i) the kinds of data intensive activities that are observed in workflows (data oriented motifs), and ii) the different manners in which activities are implemented within workflows (workflow oriented motifs). These motifs can be useful to inform workflow designers on the good and bad practices for workflow development, to inform the design of automated tools for the generation of workflow abstractions, etc. I. I NTRODUCTION Scientific workflows have been increasingly used in the last decade as an instrument for data intensive scientific analysis. In these settings, workflows serve a dual function: first as detailed documentation of the method (i. e. the input sources and processing steps taken for the derivation of a certain data item) and second as re-usable, executable artifacts for data-intensive analysis. Workflows stitch together a variety of data manipulation activities such as data movement, data transformation or data visualization to serve the goals of the scientific study. The stitching is realized by the constructs made available by the workflow system used and is largely shaped by the environment in which the system operates and the function undertaken by the workflow. A variety of workflow systems are in use [10] [3] [7] [2] serving several scientific disciplines. A workflow is a software artifact, and as such once developed and tested, it can be shared and exchanged between scientists. Other scientists can then reuse existing workflows in their experiments, e.g., as sub-workflows [17]. Workflow reuse presents several advantages [4]. For example, it enables proper data citation and improves quality through shared workflow development by leveraging the expertise of previous users. Users can also re-purpose existing workflows to adapt them to their needs [4]. Emerging workflow repositories such as myExperiment [14] and CrowdLabs [8] have made publishing and finding workflows easier, but scientists still face the challenges of reuse, which amounts to fully understanding and exploiting the available workflows/fragments. One difficulty in understanding workflows is their complex nature. A workflow may contain several scientifically-significant analysis steps, combined with various other data preparation activities, and in different implementation styles depending on the environment and context in which the workflow is executed. The difficulty in understanding causes workflow developers to revert to starting from scratch rather than re-using existing fragments. Through an analysis of the current practices in scientific workflow development, we could gain insights on the creation of understandable and more effectively re-usable workflows. Specifically, we propose an analysis with the following objectives: 1) To reverse-engineer the set of current practices in workflow development through an analysis of empirical evidence. 2) To identify workflow abstractions that would facilitate understandability and therefore effective re-use. 3) To detect potential information sources and heuristics that can be used to inform the development of tools for creating workflow abstractions. In this paper we present the result of an empirical analysis performed over 177 workflow descriptions from Taverna [10] and Wings [3]. Based on this analysis, we propose a catalogue of scientific workflow motifs. Motifs are provided through i) a characterization of the kinds of data-oriented activities that are carried out within workflows, which we refer to as dataoriented motifs, and ii) a characterization of the different manners in which those activity motifs are realized/implemented within workflows, which we refer to as workflow-oriented motifs. It is worth mentioning that, although important, motifs that have to do with scheduling and mapping of workflows onto distributed resources [12] are out the scope of this paper. The paper is structured as follows. We begin by providing related work in Section II, which is followed in Section III by brief background information on Scientific Workflows, and the two systems that were subject to our analysis. Afterwards we describe the dataset and the general approach of our analysis. We present the detected scientific workflow motifs in Section IV and we highlight the main features of their distribution
- 14. Data is the Bottleneck Common Motifs in Scientific Workflows: An Empirical Analysis Daniel Garijo⇤ , Pinar Alper † , Khalid Belhajjame† , Oscar Corcho⇤ , Yolanda Gil‡ , Carole Goble† ⇤ Ontology Engineering Group, Universidad Polit´ cnica de Madrid. {dgarijo, ocorcho}@fi.upm.es e of Computer Science, University of Manchester. {alperp, khalidb, carole.goble}@cs.manchester.ac.uk ‡ Information Sciences Institute, Department of Computer Science, University of Southern California. [email protected] † School Abstract—While workflow technology has gained momentum in the last decade as a means for specifying and enacting computational experiments in modern science, reusing and repurposing existing workflows to build new scientific experiments is still a daunting task. This is partly due to the difficulty that scientists experience when attempting to understand existing workflows, which contain several data preparation and adaptation steps in addition to the scientifically significant analysis steps. One way to tackle the understandability problem is through providing abstractions that give a high-level view of activities undertaken within workflows. As a first step towards abstractions, we report in this paper on the results of a manual analysis performed over a set of real-world scientific workflows from Taverna and Wings systems. Our analysis has resulted in a set of scientific workflow motifs that outline i) the kinds of data intensive activities that are observed in workflows (data oriented motifs), and ii) the different manners in which activities are implemented within workflows (workflow oriented motifs). These motifs can be useful to inform workflow designers on the good and bad practices for workflow development, to inform the design of automated tools for the generation of workflow abstractions, etc. I. I NTRODUCTION Scientific workflows have been increasingly used in the last decade as an instrument for data intensive scientific analysis. In these settings, workflows serve a dual function: first as detailed documentation of the method (i. e. the input sources and processing steps taken for the derivation of a certain data item) and second as re-usable, executable artifacts for data-intensive analysis. Workflows stitch together a variety of data manipulation activities such as data movement, data transformation or data visualization to serve the goals of the scientific study. The stitching is realized by the constructs made available by the workflow system used and is largely shaped by the environment in which the system operates and the function undertaken by the workflow. A variety of workflow systems are in use [10] [3] [7] [2] serving several scientific disciplines. A workflow is a software artifact, and as such once developed and tested, it can be shared and exchanged between scientists. Other scientists can then reuse existing workflows in their experiments, e.g., as sub-workflows [17]. Workflow reuse presents several advantages [4]. For example, it enables proper data citation and improves quality through shared workflow development by leveraging the expertise of previous users. Users can also re-purpose existing workflows to adapt them to their needs [4]. Emerging workflow repositories such as myExperiment [14] and CrowdLabs [8] have made publishing and finding workflows easier, but scientists still face the challenges of reuse, which amounts to fully understanding and exploiting the available workflows/fragments. One difficulty in understanding workflows is their complex nature. A workflow may contain several scientifically-significant analysis steps, combined with various other data preparation activities, and in different implementation styles depending on the environment and context in which the workflow is executed. The difficulty in understanding causes workflow developers to revert to starting from scratch rather than re-using existing fragments. Through an analysis of the current practices in scientific workflow development, we could gain insights on the creation of understandable and more effectively re-usable workflows. Specifically, we propose an analysis with the following objectives: 1) To reverse-engineer the set of current practices in workflow development through an analysis of empirical evidence. 2) To identify workflow abstractions that would facilitate understandability and therefore effective re-use. 3) To detect potential information sources and heuristics that can be used to inform the development of tools for creating workflow abstractions. In this paper we present the result of an empirical analysis performed over 177 workflow descriptions from Taverna [10] and Wings [3]. Based on this analysis, we propose a catalogue of scientific workflow motifs. Motifs are provided through i) a characterization of the kinds of data-oriented activities that are carried out within workflows, which we refer to as dataoriented motifs, and ii) a characterization of the different manners in which those activity motifs are realized/implemented within workflows, which we refer to as workflow-oriented motifs. It is worth mentioning that, although important, motifs that have to do with scheduling and mapping of workflows onto distributed resources [12] are out the scope of this paper. The paper is structured as follows. We begin by providing related work in Section II, which is followed in Section III by brief background information on Scientific Workflows, and the two systems that were subject to our analysis. Afterwards we describe the dataset and the general approach of our analysis. We present the detected scientific workflow motifs in Section IV and we highlight the main features of their distribution Data-Oriented Motifs per Domain Fig. 3. Distribution of Data-Oriented Motifs per domain
- 15. Data is the Bottleneck Common Motifs in Scientific Workflows: An Empirical Analysis Daniel Garijo⇤ , Pinar Alper † , Khalid Belhajjame† , Oscar Corcho⇤ , Yolanda Gil‡ , Carole Goble† ⇤ Ontology Engineering Group, Universidad Polit´ cnica de Madrid. {dgarijo, ocorcho}@fi.upm.es e of Computer Science, University of Manchester. {alperp, khalidb, carole.goble}@cs.manchester.ac.uk ‡ Information Sciences Institute, Department of Computer Science, University of Southern California. [email protected] † School Abstract—While workflow technology has gained momentum in the last decade as a means for specifying and enacting computational experiments in modern science, reusing and repurposing existing workflows to build new scientific experiments is still a daunting task. This is partly due to the difficulty that scientists experience when attempting to understand existing workflows, which contain several data preparation and adaptation steps in addition to the scientifically significant analysis steps. One way to tackle the understandability problem is through providing abstractions that give a high-level view of activities undertaken within workflows. As a first step towards abstractions, we report in this paper on the results of a manual analysis performed over a set of real-world scientific workflows from Taverna and Wings systems. Our analysis has resulted in a set of scientific workflow motifs that outline i) the kinds of data intensive activities that are observed in workflows (data oriented motifs), and ii) the different manners in which activities are implemented within workflows (workflow oriented motifs). These motifs can be useful to inform workflow designers on the good and bad practices for workflow development, to inform the design of automated tools for the generation of workflow abstractions, etc. Fig. 3. [14] and CrowdLabs [8] have made publishing and finding workflows easier, but scientists still face the challenges of reuse, which amounts to fully understanding and exploiting the available workflows/fragments. One difficulty in understanding workflows is their complex nature. A workflow may contain several scientifically-significant analysis steps, combined with various other data preparation activities, and in different implementation styles depending on the environment and context in which the workflow is executed. The difficulty in understanding causes workflow developers to revert to starting from scratch rather than re-using existing fragments. Through an analysis of the current practices in scientific workflow development, we could gain insights on the creation of understandable and more effectively re-usable workflows. Specifically, we propose an analysis with the following objectives: Distribution of Data-Orientedpractices in work- domain 1) To reverse-engineer the set of current Motifs per I. I NTRODUCTION Scientific workflows have been increasingly used in the last decade as an instrument for data intensive scientific analysis. In these settings, workflows serve a dual function: first as detailed documentation of the method (i. e. the input sources and processing steps taken for the derivation of a certain data item) and second as re-usable, executable artifacts for data-intensive analysis. Workflows stitch together a variety of data manipulation activities such as data movement, data transformation or data visualization to serve the goals of the scientific study. The stitching is realized by the constructs made available by the workflow system used and is largely shaped by the environment in which the system operates and the function undertaken by the workflow. A variety of workflow systems are in use [10] [3] [7] [2] serving several scientific disciplines. A workflow is a software artifact, and as such once developed and tested, it can be shared and exchanged between scientists. Other scientists can then reuse existing workflows in their experiments, e.g., as sub-workflows [17]. Workflow reuse presents several advantages [4]. For example, it enables proper data citation and improves quality through shared workflow development by leveraging the expertise of previous users. Users can also re-purpose existing workflows to adapt them to their needs [4]. Emerging workflow repositories such as myExperiment flow development through an analysis of empirical evidence. 2) To identify workflow abstractions that would facilitate understandability and therefore effective re-use. 3) To detect potential information sources and heuristics that can be used to inform the development of tools for creating workflow abstractions. In this paper we present the result of an empirical analysis performed over 177 workflow descriptions from Taverna [10] and Wings [3]. Based on this analysis, we propose a catalogue of scientific workflow motifs. Motifs are provided through i) a characterization of the kinds of data-oriented activities that are carried out within workflows, which we refer to as dataoriented motifs, and ii) a characterization of the different manners in which those activity motifs are realized/implemented within workflows, which we refer to as workflow-oriented motifs. It is worth mentioning that, although important, motifs that have to do with scheduling and mapping of workflows onto distributed resources [12] are out the scope of this paper. The paper is structured as follows. We begin by providing related work in Section II, which is followed in Section III by brief background information on Scientific Workflows, and the two systems that were subject to our analysis. Afterwards we describe the dataset and the general approach of our analysis. We present the detected scientific workflow motifs in Section IV and we highlight the main features of their distribution Fig. 5. Fig. 3. Data-Preparation Motifs per Domain Data-Oriented Motifs per Domain Data Preparation Motifs in the Genomics Wo Distribution of Data-Oriented Motifs per domain
- 16. Make Data Flourish From data to information to knowledge
- 17. Make Data Flourish From data to information to knowledge Global identification of data sets and data items Data uses a common syntax Papers explicitly link to data Metadata expressed using shared vocabularies Capture the processes by which data is manipulated Track and publish explicit provenance information
- 18. Make Data Flourish From data to information to knowledge Global identification of data sets and data items Metadata expressed using shared vocabularies Capture the processes by "Someone who is not the person who collected the data can which data is Data uses a common syntax experiment and data" - Shreejoy Tripathy manipulated understand the Papers explicitly link to data Track and publish explicit provenance information
- 19. Linked Data • • • • • Use existing Web infrastructure Everything gets a URI and usually a category Express typed relations between things (triples) Express sameness or difference Reuse identifiers as much as possible + =
- 20. Salah, Alkim Almila Akdag, Cheng Gao, Krzysztof Suchecki, and Andrea Scharnhorst. 2012. “Need to Categorize: A Comparative Look at the Categories of Universal Decimal Classification System and Wikipedia.” Leonardo 45 (1) (February): 84-85. doi:10.1162/LEON_a_00344. (Preprint http://arxiv.org/abs/1105.5912v1)
- 21. Linked Data for Science Neuroscience Information Framework (Ontologies, Semantic Wiki, Catalog) Nanopublications (small scientific assertions) Workflow Systems (WINGS, Taverna, …) Linked Science (tools) BioPortal (ontologies) Organic Data Publishing Rightfield (Semantic Wiki) (systems biology) Bio2RDF (big linked data)
- 23. Hellenic FBD Hellenic PD Crime Reports UK Ox Points NHS (EnAKTing) Ren. Energy Generators Open Election Data Project EU Institutions CO2 Emission (EnAKTing) Energy (EnAKTing) EEA Mortality (EnAKTing) Ordnance Survey legislation data.gov.uk UK Postcodes ESD standards ISTAT Immigration Lichfield Spending Scotland Pupils & Exams Traffic Scotland Data Gov.ie reference data.gov. uk London Gazette TWC LOGD Eurostat (FUB) CORDIS CORDIS (FUB) (RKB Explorer) Linked EDGAR (Ontology Central) EURES (Ontology Central) GovTrack Finnish Municipalities New York Times Italian public schools IdRef Sudoc Greek DBpedia Geo Names World Factbook Geo Species UMBEL Freebase DBLP (FU Berlin) dataopenac-uk TCM Gene DIT Daily Med SIDER Twarql EUNIS PDB SMC Journals Ocean Drilling Codices Turismo de Zaragoza Janus AMP Climbing Linked GeoData Alpine Ski Austria AEMET Metoffice Weather Forecasts Yahoo! Geo Planet National Radioactivity JP ChEMBL Open Data Thesaurus Sears DBLP (RKB Explorer) STW GESIS Budapest Pisa RESEX Scholarometer IRIT ACM NVD IBM DEPLOY Newcastle RAE2001 LOCAH Roma CiteSeer Courseware dotAC ePrints IEEE RISKS PROSITE Affymetrix SISVU GEMET Airports lobid Organisations ECS (RKB Explorer) HGNC (Bio2RDF) PubMed ProDom VIVO Cornell STITCH Linked Open Colors SGD Gene Ontology AGROV OC Product DB Weather Stations Swedish Open Cultural Heritage LAAS NSF KISTI JISC WordNet (RKB Explorer) EARTh ECS Southampton EPrints VIVO Indiana UniProt LODE WordNet (W3C) Wiki ECS Southampton Pfam LinkedCT Taxono my Cornetto NSZL Catalog P20 Eurécom totl.net WordNet (VUA) lobid Resources UN/ LOCODE Drug Bank Enipedia Lexvo DBLP (L3S) ERA Diseasome lingvoj Europeana Deutsche Biographie OAI data dcs Uberblic YAGO Open Cyc BibBase OS dbpedia lite Norwegian MeSH VIAF UB Mannheim Ulm data bnf.fr BNB Project Gutenberg Rådata nå! GND ndlna Calames DDC iServe riese GeoWord Net El Viajero Tourism URI Burner LIBRIS LCSH MARC Codes List PSH RDF Book Mashup Open Calais ntnusc Thesaurus W SW Dog Food Portuguese DBpedia LEM RAMEAU SH LinkedL CCN Sudoc UniProt US Census (rdfabout) Piedmont Accomodations Linked MDB t4gm info Open Library (Talis) theses. fr my Experiment flickr wrappr NDL subjects Plymouth Reading Lists Revyu Fishes of Texas (rdfabout) Scotland Geography Pokedex Event Media US SEC Semantic XBRL FTS Goodwin Family NTU Resource Lists Open Library SSW Thesaur us Didactal ia DBpedia Linked Sensor Data (Kno.e.sis) Eurostat Chronicling America Telegraphis Geo Linked Data Source Code Ecosystem Linked Data semantic web.org BBC Music BBC Wildlife Finder NASA (Data Incubator) transport data.gov. uk Eurostat Classical (DB Tune) Taxon Concept LOIUS Poképédia St. Andrews Resource Lists Manchester Reading Lists gnoss Last.FM (rdfize) BBC Program mes Rechtspraak. nl Openly Local data.gov.uk intervals Music Brainz (DBTune) Jamendo (DBtune) Ontos News Portal Sussex Reading Lists Bricklink yovisto Semantic Tweet Linked Crunchbase RDF ohloh (Data Incubator) (DBTune) OpenEI statistics data.gov. uk GovWILD Brazilian Politicians educatio n.data.g ov.uk Music Brainz (zitgist) Discogs FanHubz patents data.go v.uk research data.gov. uk Klappstuhlclub Lotico (Data Incubator) Last.FM artists Population (EnAKTing) reegle Surge Radio tags2con delicious Slideshare 2RDF (DBTune) Music Brainz John Peel (DBTune) EUTC Productions business data.gov. uk Crime (EnAKTing) GTAA Magnatune DB Tropes Moseley Folk Linked User Feedback LOV Audio Scrobbler OMIM MGI InterPro Smart Link Product Types Ontology Open Corporates Italian Museums Amsterdam Museum Linking Open Data cloud diagram, by Richard Cyganiak and Anja Jentzsch. http://lod-cloud.net/ UniParc UniRef UniSTS GeneID Linked Open Numbers Reactome OGOLOD KEGG Pathway Medi Care Google Art wrapper meducator KEGG Drug Pub Chem UniPath way Chem2 Bio2RDF Homolo Gene VIVO UF ECCOTCP bible ontology KEGG Enzyme PBAC KEGG Reaction KEGG Compound KEGG Glycan Media Geographic Publications User-generated content Government Cross-domain Life sciences As of September 2011
- 24. Eurostat Finnish Municipalities 0 (rdfabout) Scotland Geography US Census (rdfabout) GeoWord Net Piedmont Accomodations Italian public schools El Viajero Tourism Greek DBpedia World Factbook Geo Species UMBEL Freebase Project Gutenberg dbpedia lite DBLP (FU Berlin) dataopenac-uk TCM Gene DIT Daily Med SIDER SMC Journals Ocean Drilling Codices Turismo de Zaragoza Janus AMP EUNIS Climbing Twarql Linked GeoData WordNet (W3C) Alpine Ski Austria AEMET Metoffice Weather Forecasts WordNet (RKB Explorer) UniProt (Bio2RDF) Affymetrix SISVU GEMET ChEMBL Open Data Thesaurus Product DB Airports National Radioactivity JP LODE Taxono my Sears Linked Open Colors PDB PROSITE Open Corporates Italian Museums PubMed MGI InterPro Amsterdam Museum Linking Open Data cloud diagram, by Richard Cyganiak and Anja Jentzsch. http://lod-cloud.net/ UniRef HGNC SGD Gene Ontology OMIM UniParc UniSTS Linked Open Numbers Reactome OGOLOD Pub Chem GeneID ECS Southampton EPrints lobid Organisations ECS (RKB Explorer) DBLP (RKB Explorer) UniPath way Chem2 Bio2RDF Swedish Open Cultural Heritage STW GESIS Budapest Pisa RESEX Scholarometer IRIT ACM NVD IBM DEPLOY Newcastle RAE2001 LOCAH Roma CiteSeer Courseware KEGG Drug KEGG Pathway Homolo Gene dotAC ePrints LAAS NSF KISTI JISC VIVO UF ECCOTCP bible ontology KEGG Enzyme PBAC KEGG Reaction KEGG Compound IEEE RISKS VIVO Cornell STITCH Medi Care Google Art wrapper meducator Wiki ECS Southampton VIVO Indiana ProDom Smart Link Product Types Ontology NSZL Catalog Pfam LinkedCT AGROV OC EARTh Weather Stations Yahoo! Geo Planet Cornetto lobid Resources P20 Eurécom totl.net WordNet (VUA) Ulm UN/ LOCODE Drug Bank Enipedia Lexvo DBLP (L3S) ERA Diseasome lingvoj Europeana Deutsche Biographie OAI data dcs Uberblic YAGO Open Cyc BibBase OS VIAF UB Mannheim Calames BNB UniProt US SEC Semantic XBRL FTS Geo Names riese 8 okt. 2007 Linked EDGAR (Ontology Central) EURES (Ontology Central) GovTrack URI Burner Norwegian MeSH GND ndlna data bnf.fr iServe Fishes of Texas Linked Sensor Data (Kno.e.sis) Eurostat 1 mei 2007 CORDIS (FUB) (RKB Explorer) IdRef Sudoc DDC Open Calais Rådata nå! PSH RDF Book Mashup DBpedia Geo Linked Data CORDIS New York Times LIBRIS LCSH MARC Codes List Sudoc SW Dog Food Portuguese DBpedia ntnusc Thesaurus W 23 feb. 2012 TWC LOGD Eurostat (FUB) Event Media LEM RAMEAU SH LinkedL CCN 14 jul. 2009 Data Gov.ie 100 London Gazette NASA (Data Incubator) transport data.gov. uk Linked MDB 27 mrt. 2009 Traffic Scotland data.gov.uk intervals flickr wrappr t4gm info Open Library (Talis) theses. fr my Experiment 5 mrt. 2009 Scotland Pupils & Exams reference data.gov. uk Pokedex NDL subjects Plymouth Reading Lists Revyu Taxon Concept LOIUS Chronicling America Telegraphis 200 Goodwin Family NTU Resource Lists Open Library SSW Thesaur us semantic web.org BBC Music BBC Wildlife Finder Rechtspraak. nl Openly Local Classical (DB Tune) Source Code Ecosystem Linked Data Didactal ia 18 sep. 2008 ISTAT Immigration Lichfield Spending OpenEI statistics data.gov. uk GovWILD ESD standards educatio n.data.g ov.uk Ordnance Survey legislation data.gov.uk UK Postcodes Brazilian Politicians 300 Poképédia Last.FM (rdfize) BBC Program mes Ontos News Portal Manchester Reading Lists gnoss 31 mrt. 2008 Open Election Data Project EU Institutions CO2 Emission (EnAKTing) Energy (EnAKTing) EEA Mortality (EnAKTing) Jamendo (DBtune) 28 feb. 2008 Ren. Energy Generators (DBTune) patents data.go v.uk research data.gov. uk Music Brainz (DBTune) FanHubz Last.FM artists Population (EnAKTing) NHS (EnAKTing) (Data Incubator) yovisto Semantic Tweet Linked Crunchbase RDF ohloh Discogs 10 nov. 2007 Ox Points reegle business data.gov. uk Crime (EnAKTing) Surge Radio Music Brainz (zitgist) (Data Incubator) 7 nov. 2007 Crime Reports UK 400 Lotico St. Andrews Resource Lists 19 sep. 2011 Hellenic PD EUTC Productions Klappstuhlclub Sussex Reading Lists Bricklink (DBTune) Music Brainz John Peel (DBTune) tags2con delicious Slideshare 2RDF 22 sep. 2010 Hellenic FBD GTAA Magnatune DB Tropes Moseley Folk Linked User Feedback LOV Audio Scrobbler KEGG Glycan Media Geographic Publications User-generated content Government Cross-domain Life sciences As of September 2011
- 27. LODStats Analysis http://stats.lod2.eu 140 134 2% 4% 4% 105 7% 84 35 HTTP Other 12 Unknown response 28% 30 No URL provided 11 XML 6 Connection reset 0 Not RDF 22 10% 45% 70 Not RDF Connection reset Unknown response XML No URL provided Other HTTP Hoekstra, Rinke; Groth, Paul (2013): Distribution of Errors Reported by LOD2 LODStats Project. figshare. http://dx.doi.org/10.6084/m9.figshare.695949 299 out of 639 datasets have errors
- 28. An Ambient Agent Model for Monitoring and Analysing Dynamics of Complex Human Behaviour Tibor Bossea*, Mark Hoogendoorna, Michel C.A. Kleina, and Jan Treura a Vrije Universiteit Amsterdam, Department of Artificial Intelligence, de Boelelaan 1081, 1081 HV Amsterdam, The Netherlands Abstract. In ambient intelligent systems, monitoring of a human could consist of more complex tasks than merely identifying whether a certain value of a sensor is above a certain threshold. Instead, such tasks may involve monitoring of complex dynamic interactions between human and environment. In order to enable such more complex types of monitoring, this paper presents a generic agent-based framework. The framework consists of support on various levels of system design, namely: (1) the top level, including the interaction between agents, (2) the agent level, providing support on the design of individual agents, and (3) the level of monitoring complex dynamic behaviour, allowing the specification of the aforementioned complex monitoring properties within the agents. The approach is exemplified by a large case study concerning the assessment of driving behaviour, and is applied to two smaller cases as well (concerning fall detection of elderly, and assistance of naval operations, respectively), which are briefly described. These case studies have illustrated that the presented framework enables developers within ambient intelligence to build systems with more expressiveness regarding their monitoring focus. Moreover, they have shown that the framework is easy to use and applicable in a wide variety of domains. Keywords: ambient agent model, human behaviour, dynamics Journal of Ambient Intelligence and Smart Environments
- 29. “Whoah! Cool, you should publish that stuff as Linked Data” An Ambient Agent Model for Monitoring and Analysing Dynamics of Complex Human Behaviour Tibor Bossea*, Mark Hoogendoorna, Michel C.A. Kleina, and Jan Treura a Vrije Universiteit Amsterdam, Department of Artificial Intelligence, de Boelelaan 1081, 1081 HV Amsterdam, The Netherlands Abstract. In ambient intelligent systems, monitoring of a human could consist of more complex tasks than merely identifying whether a certain value of a sensor is above a certain threshold. Instead, such tasks may involve monitoring of complex dynamic interactions between human and environment. In order to enable such more complex types of monitoring, this paper presents a generic agent-based framework. The framework consists of support on various levels of system design, namely: (1) the top level, including the interaction between agents, (2) the agent level, providing support on the design of individual agents, and (3) the level of monitoring complex dynamic behaviour, allowing the specification of the aforementioned complex monitoring properties within the agents. The approach is exemplified by a large case study concerning the assessment of driving behaviour, and is applied to two smaller cases as well (concerning fall detection of elderly, and assistance of naval operations, respectively), which are briefly described. These case studies have illustrated that the presented framework enables developers within ambient intelligence to build systems with more expressiveness regarding their monitoring focus. Moreover, they have shown that the framework is easy to use and applicable in a wide variety of domains. Keywords: ambient agent model, human behaviour, dynamics Journal of Ambient Intelligence and Smart Environments
- 30. “Whoah! Cool, you should publish that stuff as Linked Data” An Ambient Agent Model “Um, but doesn’t TTL have incompatible semantics?” for Monitoring and Analysing Dynamics of Complex Human Behaviour Tibor Bossea*, Mark Hoogendoorna, Michel C.A. Kleina, and Jan Treura a Vrije Universiteit Amsterdam, Department of Artificial Intelligence, de Boelelaan 1081, 1081 HV Amsterdam, The Netherlands Abstract. In ambient intelligent systems, monitoring of a human could consist of more complex tasks than merely identifying whether a certain value of a sensor is above a certain threshold. Instead, such tasks may involve monitoring of complex dynamic interactions between human and environment. In order to enable such more complex types of monitoring, this paper presents a generic agent-based framework. The framework consists of support on various levels of system design, namely: (1) the top level, including the interaction between agents, (2) the agent level, providing support on the design of individual agents, and (3) the level of monitoring complex dynamic behaviour, allowing the specification of the aforementioned complex monitoring properties within the agents. The approach is exemplified by a large case study concerning the assessment of driving behaviour, and is applied to two smaller cases as well (concerning fall detection of elderly, and assistance of naval operations, respectively), which are briefly described. These case studies have illustrated that the presented framework enables developers within ambient intelligence to build systems with more expressiveness regarding their monitoring focus. Moreover, they have shown that the framework is easy to use and applicable in a wide variety of domains. Keywords: ambient agent model, human behaviour, dynamics Journal of Ambient Intelligence and Smart Environments
- 31. “Whoah! Cool, you should publish that stuff as Linked Data” An Ambient Agent Model “Um, but doesn’t TTL have incompatible semantics?” for Monitoring and Analysing Dynamics of Complex Human “Nah, silly, who cares? We’ll just start a new W3C WG!” Behaviour Tibor Bossea*, Mark Hoogendoorna, Michel C.A. Kleina, and Jan Treura a Vrije Universiteit Amsterdam, Department of Artificial Intelligence, de Boelelaan 1081, 1081 HV Amsterdam, The Netherlands Abstract. In ambient intelligent systems, monitoring of a human could consist of more complex tasks than merely identifying whether a certain value of a sensor is above a certain threshold. Instead, such tasks may involve monitoring of complex dynamic interactions between human and environment. In order to enable such more complex types of monitoring, this paper presents a generic agent-based framework. The framework consists of support on various levels of system design, namely: (1) the top level, including the interaction between agents, (2) the agent level, providing support on the design of individual agents, and (3) the level of monitoring complex dynamic behaviour, allowing the specification of the aforementioned complex monitoring properties within the agents. The approach is exemplified by a large case study concerning the assessment of driving behaviour, and is applied to two smaller cases as well (concerning fall detection of elderly, and assistance of naval operations, respectively), which are briefly described. These case studies have illustrated that the presented framework enables developers within ambient intelligence to build systems with more expressiveness regarding their monitoring focus. Moreover, they have shown that the framework is easy to use and applicable in a wide variety of domains. Keywords: ambient agent model, human behaviour, dynamics Journal of Ambient Intelligence and Smart Environments
- 32. “Whoah! Cool, you should publish that stuff as Linked Data” An Ambient Agent Model “Um, but doesn’t TTL have incompatible semantics?” for Monitoring and Analysing Dynamics of Complex Human “Nah, silly, who cares? We’ll just start a new W3C WG!” Behaviour “Uh, ok, if we must. But, Mark Hoogendoorn , Michel C.A. Klein , and Jan Treur even then, we can’t just publish the model as is!” Tibor Bosse a* a a a a Vrije Universiteit Amsterdam, Department of Artificial Intelligence, de Boelelaan 1081, 1081 HV Amsterdam, The Netherlands Abstract. In ambient intelligent systems, monitoring of a human could consist of more complex tasks than merely identifying whether a certain value of a sensor is above a certain threshold. Instead, such tasks may involve monitoring of complex dynamic interactions between human and environment. In order to enable such more complex types of monitoring, this paper presents a generic agent-based framework. The framework consists of support on various levels of system design, namely: (1) the top level, including the interaction between agents, (2) the agent level, providing support on the design of individual agents, and (3) the level of monitoring complex dynamic behaviour, allowing the specification of the aforementioned complex monitoring properties within the agents. The approach is exemplified by a large case study concerning the assessment of driving behaviour, and is applied to two smaller cases as well (concerning fall detection of elderly, and assistance of naval operations, respectively), which are briefly described. These case studies have illustrated that the presented framework enables developers within ambient intelligence to build systems with more expressiveness regarding their monitoring focus. Moreover, they have shown that the framework is easy to use and applicable in a wide variety of domains. Keywords: ambient agent model, human behaviour, dynamics Journal of Ambient Intelligence and Smart Environments
- 33. “Whoah! Cool, you should publish that stuff as Linked Data” An Ambient Agent Model “Um, but doesn’t TTL have incompatible semantics?” for Monitoring and Analysing Dynamics of Complex Human “Nah, silly, who cares? We’ll just start a new W3C WG!” Behaviour “Uh, ok, if we must. But, Mark Hoogendoorn , Michel C.A. Klein , and Jan Treur even then, we can’t just publish the model as is!” Tibor Bosse a* a a a a Vrije Universiteit Amsterdam, Department of Artificial Intelligence, de Boelelaan 1081, 1081 HV Amsterdam, The Netherlands “”No worries, just add the provenance using PROV-O, annotate the PDF Abstract. In ambient intelligent systems, monitoring of a human could consist of more link to other research using CITO.” with OA, tasks may involvetasks than merelycomplex dyand complex monitoring of identifying whether a certain value of a sensor is above a certain threshold. Instead, such namic interactions between human and environment. In order to enable such more complex types of monitoring, this paper presents a generic agent-based framework. The framework consists of support on various levels of system design, namely: (1) the top level, including the interaction between agents, (2) the agent level, providing support on the design of individual agents, and (3) the level of monitoring complex dynamic behaviour, allowing the specification of the aforementioned complex monitoring properties within the agents. The approach is exemplified by a large case study concerning the assessment of driving behaviour, and is applied to two smaller cases as well (concerning fall detection of elderly, and assistance of naval operations, respectively), which are briefly described. These case studies have illustrated that the presented framework enables developers within ambient intelligence to build systems with more expressiveness regarding their monitoring focus. Moreover, they have shown that the framework is easy to use and applicable in a wide variety of domains. Keywords: ambient agent model, human behaviour, dynamics Journal of Ambient Intelligence and Smart Environments
- 34. “Whoah! Cool, you should publish that stuff as Linked Data” An Ambient Agent Model “Um, but doesn’t TTL have incompatible semantics?” for Monitoring and Analysing Dynamics of Complex Human “Nah, silly, who cares? We’ll just start a new W3C WG!” Behaviour “Uh, ok, if we must. But, Mark Hoogendoorn , Michel C.A. Klein , and Jan Treur even then, we can’t just publish the model as is!” Tibor Bosse a* a a a a Vrije Universiteit Amsterdam, Department of Artificial Intelligence, de Boelelaan 1081, 1081 HV Amsterdam, The Netherlands “”No worries, just add the provenance using PROV-O, annotate the PDF Abstract. In ambient intelligent systems, monitoring of a human could consist of more link to other research using CITO.” with OA, tasks may involvetasks than merelycomplex dyand complex monitoring of identifying whether a certain value of a sensor is above a certain threshold. Instead, such namic interactions between human and environment. In order to enable such more complex types of monitoring, this paper “And that’s it?” a generic agent-based framework. The framework consists of support on various levels of system design, namely: (1) presents the top level, including the interaction between agents, (2) the agent level, providing support on the design of individual agents, and (3) the level of monitoring complex dynamic behaviour, allowing the specification of the aforementioned complex monitoring properties within the agents. The approach is exemplified by a large case study concerning the assessment of driving behaviour, and is applied to two smaller cases as well (concerning fall detection of elderly, and assistance of naval operations, respectively), which are briefly described. These case studies have illustrated that the presented framework enables developers within ambient intelligence to build systems with more expressiveness regarding their monitoring focus. Moreover, they have shown that the framework is easy to use and applicable in a wide variety of domains. Keywords: ambient agent model, human behaviour, dynamics Journal of Ambient Intelligence and Smart Environments
- 35. “Whoah! Cool, you should publish that stuff as Linked Data” An Ambient Agent Model “Um, but doesn’t TTL have incompatible semantics?” for Monitoring and Analysing Dynamics of Complex Human “Nah, silly, who cares? We’ll just start a new W3C WG!” Behaviour “Uh, ok, if we must. But, Mark Hoogendoorn , Michel C.A. Klein , and Jan Treur even then, we can’t just publish the model as is!” Tibor Bosse a* a a a a Vrije Universiteit Amsterdam, Department of Artificial Intelligence, de Boelelaan 1081, 1081 HV Amsterdam, The Netherlands “”No worries, just add the provenance using PROV-O, annotate the PDF Abstract. In ambient intelligent systems, monitoring of a human could consist of more link to other research using CITO.” with OA, tasks may involvetasks than merelycomplex dyand complex monitoring of identifying whether a certain value of a sensor is above a certain threshold. Instead, such namic interactions between human and environment. In order to enable such more complex types of monitoring, this paper “And that’s it?” a generic agent-based framework. The framework consists of support on various levels of system design, namely: (1) presents the top level, including the interaction between agents, (2) the agent level, providing support on the design of individual agents, and (3) the level of monitoring complex dynamic behaviour, allowing the specification of the aforementioned complex moni“Noo! You’ll need persistent Cool URI’s and publish your endpoint toring properties within the agents. The approach is exemplified by a large case study concerning the assessment of driving behaviour, and is applied to two smaller cases as well (concerning fall detection of elderly, and assistance of naval operations, respectively), which are briefly described. These case studies have illustrated that the presented framework enables developers for eternity of course. Duh.” within ambient intelligence to build systems with more expressiveness regarding their monitoring focus. Moreover, they have shown that the framework is easy to use and applicable in a wide variety of domains. Keywords: ambient agent model, human behaviour, dynamics Journal of Ambient Intelligence and Smart Environments
- 36. “Whoah! Cool, you should publish that stuff as Linked Data” An Ambient Agent Model “Um, but doesn’t TTL have incompatible semantics?” for Monitoring and Analysing Dynamics of Complex Human “Nah, silly, who cares? We’ll just start a new W3C WG!” Behaviour “Uh, ok, if we must. But, Mark Hoogendoorn , Michel C.A. Klein , and Jan Treur even then, we can’t just publish the model as is!” Tibor Bosse a* a a a a Vrije Universiteit Amsterdam, Department of Artificial Intelligence, de Boelelaan 1081, 1081 HV Amsterdam, The Netherlands “”No worries, just add the provenance using PROV-O, annotate the PDF Abstract. In ambient intelligent systems, monitoring of a human could consist of more link to other research using CITO.” with OA, tasks may involvetasks than merelycomplex dyand complex monitoring of identifying whether a certain value of a sensor is above a certain threshold. Instead, such namic interactions between human and environment. In order to enable such more complex types of monitoring, this paper “And that’s it?” a generic agent-based framework. The framework consists of support on various levels of system design, namely: (1) presents the top level, including the interaction between agents, (2) the agent level, providing support on the design of individual agents, and (3) the level of monitoring complex dynamic behaviour, allowing the specification of the aforementioned complex moni“Noo! You’ll need persistent Cool URI’s and publish your endpoint toring properties within the agents. The approach is exemplified by a large case study concerning the assessment of driving behaviour, and is applied to two smaller cases as well (concerning fall detection of elderly, and assistance of naval operations, respectively), which are briefly described. These case studies have illustrated that the presented framework enables developers for eternity of course. Duh.” within ambient intelligence to build systems with more expressiveness regarding their monitoring focus. Moreover, they have shown that the framework is easy to use and applicable in a wide variety of domains. “Eh?” Keywords: ambient agent model, human behaviour, dynamics “Oh... and don’t forget all data collected by the agents, in all runs, including the first experiments. Now THAT would be ultra cool. “Ngh!?” Journal of Ambient Intelligence and Smart Environments
- 37. Creating Linked Data http://linkeddatabook.com • • • • • • • • Decide on resources to describe Mint cool URIs Decide on triples to include Describe the dataset Choose vocabularies Define terms Make links Publish to triple store/annotations/dump
- 38. If this already is tedious... ... can you expect researchers to publish Linked Research Data?
- 39. If this already is tedious... ... can you expect researchers to publish Linked Research Data?
- 40. Conclusion?
- 41. We need to make publishing Linked Research Data... ...a lot easier... ... more persistent ... ... and more rewarding. Linked Data is sóóóóó 2005
- 42. We need to make publishing Linked Research Data... ...a lot easier... ... more persistent ... ... and more rewarding. “People as frontier in computing” - Haym Hirsch, Pietro Michelucci
- 43. We need to make publishing Linked Research Data... ...a lot easier... ... more persistent ... ... and more rewarding. http://linkitup.data2semantics.org
- 44. We need to make publishing Linked Research Data... ...a lot easier... ... more persistent ... • • • • • • ... and more rewarding. Lightweight web application Interface to API of existing data repositories Enrich metadata by linking to (linked) data resources Human in the Loop Track provenance Publish rich metadata as new data publication Nanopublication + OA + PROV-O + DCTerms + FOAF http://linkitup.data2semantics.org
- 45. We need to make publishing Linked Research Data... ...a lot easier... ... more persistent ... • • • • • • ... and more rewarding. Lightweight web application Interface to API of existing data repositories Enrich metadata by linking to (linked) data resources Human in the Loop Track provenance Publish rich metadata as new data publication Nanopublication + OA + PROV-O + DCTerms + FOAF http://linkitup.data2semantics.org
- 49. Use tags & categories to query the DBpedia endpoint
- 50. Use authors to query the DBLP endpoint
- 51. Use tags & categories to query the NeuroLex endpoint
- 52. Use author names to query the ORCID API
- 53. Extract references to resolve to CrossRef DOIs
- 54. Every operation is tracked automatically
- 56. Review selected links, and publish to Figshare
- 58. Plugins Name DBLP ORCID LinkedLifeData Crossref Elsevier LDR DANS EASY SameAs DBPedia Spotlight DBPedia/Wikipedia NeuroLex NIF Registry your Service SPARQL REST REST Custom REST Custom REST REST SPARQL SPARQL REST data Source Authors Authors Tags & Categories Citations Tags & Categories Tags & Categories Links Description, Tags & Categories Tags & Categories Tags & Categories Tags & Categories set Links to Author Identifiers Author Identifiers Biomedical Entities DOIs Funding agencies General Datasets General Entities General Entities General Entities Neuroscience Concepts Neuroscience Datasets here
- 59. What does this solve? http://linkeddatabook.com • • • • • • • • Decide on resources to describe Mint cool URIs Decide on triples to include Describe the dataset Choose vocabularies Define terms Make links Publish to triple store/annotations/dump
- 60. What does this solve? http://linkeddatabook.com • • • • • • • • You decide on resources to describe We mint cool URIs We decide on triples to include We describe the dataset We choose vocabularies We define terms Together we make links We publish the dataset to a reliable repository
- 61. Coming up… • • • • • • Publish directly from Dropbox, Github, … Reconstruct provenance information (http://git2prov.org) Analyze, convert and enrich on the fly Generate a data report for advertisement purposes Measure for information content of datasets (“D-Index”) Integrate a data dashboard
- 62. 84 70 12 22 30 HTTP 11 Other 6 No URL provided 0 XML 35 Unknown response … enhancing the data publication… 105 Connection reset http://linkitup.data2semantics.org 134 Not RDF linkitup 140 … increasing findability … … boosting reusability … … result is stored persistently http://git2prov.org http://semweb.cs.vu.nl/provoviz http://yasgui.data2semantics.org http://www.data2semantics.org