Primo and the Semantic Web Dominique Ritze Mannheim University Library.

Primo and the

Semantic Web

Dominique RitzeMannheim University Library

Linked [Open] Data Cloud

Which?

How?Why?

Overview

• Why?

• How?• Client-side enrichment• Server-side enrichment• Realization in Primo

• Which data?• Examples in Primo

Why use LD?

• Providing additional information• Wikipedia articles• External links

• Multilingual semantic search• No language barriers• No knowledge about technical terms necessary• “intelligent“ search

Enrichment

Client-side Enrichment

• Pro:• Provide additional information• Minimal-invasive solution• On-the-fly presentation• Load on the client

• Contra:• Load on the client • Data not searchable• No control over the data

Client-side Enrichment

• Examples: • Additional links to resources, e.g. Wikipedia

• Implementation• Query service to get further informationen, e.g. by

using SPARQL

• Integrate information into the website, e.g. by using JavaScript

Server-side Enrichment

• Pro:• Searchable data• Data is integrated with other data in the system• Load on the server

• Contra:• Load on the server• Replication of the data• No current data• Problems with licences

Server-side Enrichment

• Example:• Classification information

• Possibility to refine the search

• Implementation:• Loading the data into the database

• Additional information are part of the website

Discovery Systems & LD

• No discovery system provides general LD support• Workarounds

• Requirements:• Identification of resources (URI)• Access mechanisms for LD• Plugin mechanisms• Proper presentation

Primo Plugin-API

• Example:

Primo Plugin-API

• Client-side enrichment using JavaScript

• Allows to develop JavaScript plugins for Primo

• Abstraction from Primo

• Plugins can be used to enrich Primo

Primo Plugin-API

PermaLink

• Creates permanent link (URI) per resource unique identification, cool URIs [1]

• First part of the link: http://link.bib.uni-mannheim.de/primo/

• Second part of the link:Resource ID

[1] http://www.w3.org/TR/cooluris/

PermaLink Example

Database Recommender

• Recommends databases according to the search terms

Better and easier research

• Uses a web service (SuUB Bremen) to assign subjects• Based on linguistic and statistic analysis• Assignment subject – database performed by

subject specialists

Database Recommender

• Recommend databases for the entered keywords

Wikipedia Article

• Shows Wikipedia articles of authors Additional information

• First paragraph is directly shown, further link to the articles itself

• Uses assignment author – name authority file• Linkresolver to get the article• Next step: DBPedia

Wikipedia Article

• Displays Wikipedia articles for authors

Wikipedia Article

Idea: Connecting publications and research data

• On which data does a publication base on?

• Links between research data and publications are available (InFoLiS-project)

Additional information, facilitates research

Why LD support?

• Need some services to provide the data• Need to query the service• Need to parse the different formats• Need an implementation for each plugin

• LD provides the data in a standardized way• Offers new possibilities, saves a lot of time

and effort

Conclusion

• LD can be useful for libraries• Client- and server-side enrichment• feasability? searchability? replication?

• Primo Plugin-API client-side solution• PermaLink, Wikipedia Article, Database

Recommender

• Additional effort without LD support

Thank you for your attention!

[email protected]