Semantically Mapping Science (SMS) Platform · Towards an Open Infrastructure for Studying Science,...
1
Towards an Open Infrastructure for Studying Science, Technology and Innovation Semantically Mapping Science (SMS) Platform Linked Data Creation Linked Data Services Applications Use Cases Data Ingestion RISIS Public Data RISIS Private Data Open Data on the Web Adaptive Functional Urban Areas (FUAs) to Study Innovative Activities Gendered Dimensions in Grant Selection http://sms.risis.eu combine one or more SMS services with other existing services and applications to build novel and innovative applications. a b c d a) Dataset Metadata Editor b) RISIS Datasets Portal c) Spreadsheet add-on d) Address to Boundaries In order to enable batch processing of data, SMS provides a Google spreadsheet add-on. This add-on allows users to enrich their data directly in their spreadsheets. Given an address, this application returns the geographical boundaries from different open datasets. The results are displayed on a map. The RISIS dataset holders all have to describe their datasets in a detailed, consistent, and uniform way. To achieve this goal, RDF data model is used to describe the RISIS datasets. To stimulate non-Semantic Web users to generate valid RDF metadata descriptions, we designed a novel user-friendly editor which hides the complexity of RDF from non-technical users. The metadata editor exploits the state of the art Web technologies to provide user-friendly component to view and edit the metadata. In order to exploit the generated metadata, RISIS datasets portal brings a user interface to view and browse the metadata. Faceted browsing allows users to explore the dataset via multiple entry points, or when users do not know what they are looking for beforehand. The portal also handles user registration and supports the process of reviewing visit/access requests to certain RISIS datasets. In this use case, we investigated whether gender of applicants influences the grant decision. In order to answer that question, one needs a multitude of variables that may influence the grant decision - as suggested by various theories: variables representing merit (such as scholarly performance), but also variables bring in bias, such as personal characteristics (gender, age, nationality, university of PhD degree). And some in-between variables, such as the quality of the applicant's network. We use the SMS platform for preprocessing, for converting into RDF, for entity recognition, and for linking data. These variables come from a variety of data sources: o Bibliometric performance scores: Web of Science (TXT) o Quality of the applicant's network: o Organizations mentioned in the CV (PDF) o Ranking of those organizations from Leiden Ranking (Excel) o Earlier grants: from CV (PDF) o Host institution from admin file (Excel) o Other available open datasets In this use case, we investigated the effects of socio-economic and structural properties of urban areas on the level of innovative activities, as stimulated by recent research and innovation policies in the Netherlands. The funded projects can be considered as a useful representation of RTD (Research Technology Development) collaboration for innovation. In this case we are interested in the geographical properties of these projects. In order to investigate this, one needs data about the projects and data about the characteristics of the relevant geographical units. These data are available as open data. In this case, the following open datasets are deployed: o RVO dataset providing a list of research and innovation projects that have received subsidies and financial support from the Netherlands Enterprise Agency (Rijksdienst voor Ondernemend Nederland). Projects information includes companies and research institutes that are collaborating in the project, but also the geographical coordinates of the project main applicant. o CBS dataset published by the Statistics Netherlands provides different types of statistical information on dimensions such as labor and income, economy, society and regional aspects of municipalities and regions in the Netherlands. o Open Datasets on administrative boundaries such as OpenStreetMap & GADM. Browsing Cordis H2020 Research Grants Amount of RVO project subsidies mapped to the dynamically delineated FUAs defined based on the CBS open statistical data and OpenStreetMap boundaries. Integrating multiple open datasets about geographical boundaries with existing statistical data on the Web. convert unstructured and structured data to RDF. This poster is designed by Ali Khalili <http://ali1k.com> Peter van den Besselaar # , Ali Khalili*, Klaas Andries de Graaf*, Al Idrissou # , Antonis Loizou*, Stefan Schlobach* and Frank van Harmelen* Vrije Universiteit Amsterdam, The Network Institute, # Department of Organization Sciences, * Department of Computer Science expose the functionality of the SMS platform to third-party users by standard Application Programming Interfaces (APIs). Identity Resolution Services Named Entity Recognition Services Metadata Services Category Services Basic Geo Services Innovative Geo Services Integration Services handle the entity disambiguation problem and manage the same or similar entities found in different datasets. extract and classify named entities (e.g. people, places, organizations) in unstructured text. resolve the mapping between heterogeneous classification schemas (e.g. WoS and FoS, or different ISO codes for countries). allow search on datasets based on the description of datasets in multiple metadata categories (e.g. language, time coverage, etc.). deal with basic representation of geographical data together with geocoding functions and identifying the boundaries containing a given point. provide innovative services based on new notions of distance (e.g. traffic congestion, language factor, flight routes, etc.) allow integration of data with RISIS public and private datasets as well as open datasets and social data available on the Web. DBpedia Wikidata … OrgRef GRID FundRef Geoname ISNI VIAF Cordis ? Scientific Lenses When comparing two entities, depending on the user’s perspective and the context of study, they might be considered the same, similar or different. Sometimes two organizations (e.g. departments) can be the same – because they are parts of the same organization (university). But if one wants to compare departments, this is not the case. Scientific lenses support linking the entities based on the context of use and provided features of data. access and retrieve heterogeneous data. Access Control Points (ACPs) - provide standard interfaces to reduce technical difficulties of accessing data. - provide a mechanism to coordinate access to data based on the user role and the datasets owner’s requirements. WWW e e) Faceted browser Allows browsing a dataset using a set of pre-defined facets
Transcript of Semantically Mapping Science (SMS) Platform · Towards an Open Infrastructure for Studying Science,...
Towards an Open Infrastructure for Studying Science, Technology and InnovationSemantically Mapping Science (SMS) Platform
Linked Data Creation
Linked Data Services
Applications
Use Cases
Data Ingestion
RIS
IS P
ublic
Dat
a
RIS
IS P
rivat
e D
ata
Ope
n D
ata
on th
e W
eb
Adaptive Functional Urban Areas (FUAs) to Study Innovative ActivitiesGendered Dimensions in Grant Selection
http://sms.risis.eu
combineoneormoreSMSserviceswithotherexistingservicesandapplicationstobuildnovelandinnovativeapplications.
a
b
c
d
a) Dataset Metadata Editor
b) RISIS Datasets Portal
c) Spreadsheet add-on
d) Address to Boundaries
In order to enable batch processing of data, SMS provides a Google spreadsheet add-on. This add-on allows users to enrich their data directly in their spreadsheets.
Given an address, this application returns the geographical boundaries from different open datasets. The results are displayed on a map.
The RISIS dataset holders all have to describe their datasets in a detailed, consistent, and uniform way.
To achieve this goal, RDF data model is used to describe the RISIS datasets. To stimulate non-Semantic Web users to generate valid RDF metadata descriptions, we designed a novel user-friendly editor which hides the
complexity of RDF from non-technical users. The metadata editor exploits the state of the art Web technologies to provide user-friendly component to view and edit the metadata.
In order to exploit the generated metadata, RISIS datasets portal brings a user interface to view and browse the metadata. Facetedbrowsingallowsuserstoexplorethedatasetviamultipleentrypoints,orwhenusersdonotknowwhattheyarelookingforbeforehand. The portal also handles user registration and supports the process of reviewing visit/access requests to certain
RISIS datasets.
In this use case, we investigated whether gender of applicants influences the grant decision.In order to answer that question, one needs a multitude of variables that may influence thegrant decision - as suggested by various theories: variables representing merit (such asscholarly performance), but also variables bring in bias, such as personal characteristics(gender, age, nationality, university of PhD degree). And some in-between variables, such asthe quality of the applicant's network. We use the SMS platform for preprocessing, forconverting into RDF, for entity recognition, and for linking data. These variables come from avariety of data sources:o Bibliometric performance scores: Web of Science (TXT)o Quality of the applicant's network:
o Organizations mentioned in the CV (PDF)o Ranking of those organizations from Leiden Ranking (Excel)
o Earlier grants: from CV (PDF)o Host institution from admin file (Excel)o Other available open datasets
In this use case, we investigated the effects of socio-economic and structural properties of urbanareas on the level of innovative activities, as stimulated by recent research and innovation policies inthe Netherlands. The funded projects can be considered as a useful representation of RTD (ResearchTechnology Development) collaboration for innovation. In this case we are interested in thegeographical properties of these projects. In order to investigate this, one needs data about theprojects and data about the characteristics of the relevant geographical units. These data are availableas open data. In this case, the following open datasets are deployed:o RVO dataset providing a list of research and innovation projects that have received subsidies and
financial support from the Netherlands Enterprise Agency (Rijksdienst voor OndernemendNederland). Projects information includes companies and research institutes that are collaboratingin the project, but also the geographical coordinates of the project main applicant.
o CBS dataset published by the Statistics Netherlands provides differenttypes of statistical information on dimensions such as labor and income,economy, society and regional aspects of municipalities and regions in theNetherlands.
o Open Datasets on administrative boundaries such as OpenStreetMap & GADM.
Browsing Cordis H2020 Research Grants Amount of RVO project subsidies mapped to the dynamically delineated FUAs defined based on the CBS open statistical data and OpenStreetMap boundaries.
Integrating multiple open datasets about geographical boundaries with existing statistical data on the Web.
convertunstructuredandstructureddatatoRDF.
ThisposterisdesignedbyAliKhalili <http://ali1k.com>
Peter van den Besselaar#, Ali Khalili*, Klaas Andries de Graaf*, Al Idrissou#, Antonis Loizou*, Stefan Schlobach* and Frank van Harmelen*Vrije Universiteit Amsterdam, The Network Institute,
# Department of Organization Sciences, * Department of Computer Science
exposethefunctionalityoftheSMSplatformtothird-partyusersbystandardApplicationProgrammingInterfaces(APIs).
IdentityResolutionServices
NamedEntityRecognitionServices
MetadataServices
CategoryServices
BasicGeoServices
InnovativeGeoServices
IntegrationServices
handle the entity disambiguation problem and manage the same or similar entities found in different datasets.
extract and classify named entities (e.g. people, places, organizations) in unstructured text.
resolve the mapping between heterogeneous classification schemas (e.g. WoS and FoS, or different ISO codes for countries).
allow search on datasets based on the description of datasets in multiple metadata categories (e.g. language, time coverage, etc.).
deal with basic representation of geographical data together with geocoding functions and identifying the boundaries containing a given point.
provide innovative services based on new notions of distance (e.g. traffic congestion, language factor, flight routes, etc.)
allow integration of data with RISIS public and private datasets as well as open datasets and social data available on the Web.
DBpedia
Wikidata
…
OrgRef
GRID FundRef
Geoname
ISNI
VIAF
Cordis
?
Scientific LensesWhencomparingtwoentities,dependingontheuser’sperspectiveandthecontextofstudy,theymightbeconsideredthesame,similarordifferent.Sometimestwoorganizations(e.g.departments)canbethesame– becausetheyarepartsofthesameorganization(university).Butifonewantstocomparedepartments,thisisnotthecase.Scientificlensessupportlinkingtheentitiesbasedonthecontextofuseandprovidedfeaturesofdata.
accessandretrieveheterogeneousdata.
Access Control Points (ACPs)- providestandardinterfacestoreduce
technicaldifficultiesofaccessingdata.- provideamechanismtocoordinateaccess
todatabasedontheuserroleandthedatasetsowner’srequirements.
WWW
e
e) Faceted browserAllows browsing a dataset using a set of pre-defined facets
