ls Frequently Asked Questions
The Semantic Web Search Engine

Frequently Asked Questions

General

Q: What does SWSE stand for?

Semantic Web Search Engine. People pronounce it "swuasziee", "swishy", or "s-w-s-e". We prefer "swizzy".

Q: What does SWSE do?

SWSE is a search engine for the RDF Web on the Web, and provides the equivalent services a search engine currently provides for the HTML Web. The system explores and indexes the Semantic Web and provides an easy-to-use interface through which users can find the information they are looking for. Because of the inherent semantics of RDF and other Semantic Web languages, the search and information retrieval capabilities of SWSE are potentially much more powerful than those of current search engines.

Q: Who is behind SWSE?

SWSE is a research project being carried out by DERI Galway. Contributors include:

Q: What kind of data does SWSE index?

SWSE indexes RDF data which is retrieved from many sources. The main sources are OWL, RDF and RSS files. RSS2 is converted to RDF. We will add GRDDL sources soon.

Q: How often do you update the index?

We are aiming for weekly crawls. The latest dataset is from the week of 2008-03-31.

How-To

Q: How to search for general information?

SWSE's default search mode will turn up results with any of the words in your query. To search for information using SWSE

  1. Enter your keyword in the search box e.g. "social networking"
  2. Scroll through these to find instances relevant to you, or
  3. Select a class (sidebar), e.g. Post
  4. Scroll through these to find instances relevant to you

Once you have selected an entity, you can select an incoming or outgoing link at the sidebar. You then get a list of entities associated with the current entity.

Q: Do you offer SPARQL query facilities?

The endpoint supporting core SPARQL queries is available at http://swse.deri.org/yars2/.

The SPARQL endpoint is experimental. Please let us know about problems you might encounter.

Q: Who is using the SPARQL query facilities?

How to submit URIs to SWSE?

SWSE offers a service to submit your URIs as Sitemaps. You can submit your sitemap in two different ways,

Submitting URIs using an HTTP GET request Submitting URIs using the Web Interface
  1. Visit the SWSE Sitemap submission interface
  2. Enter your sitemap URI in the provided text field
  3. Click submit
  4. Wait for the status message for your submission
SWSE includes your submitted URIs in the next index build (at the moment weekly new builds).

Q: I have a FOAF file, but I'm not in your index!?

We try to get link rel links from HTML pages to FOAF files, but we might have missed your HTML homepage. The safe way is to get a rdfs:seeAlso link from somebody's FOAF file that is already in the index, or alternatively, you can add your file to the FOAF Bulletin Board page which is included in our crawl.

Q: My FOAF file is in your index, but there are statements missing!?

We normalise the data we put into our index, and sometimes statements get removed due to syntax errors. Please make sure your URIs do not contain illegal characters. One of the more common case of illegal URIs are those with spaces. Escape these characters if you really need them. We handle Unicode URIs.

Implementation

Q: What browsers does SWSE support?

Our pages are XHTML 1.0/CSS2 compliant, using client-side XSLT. The pages should display on any modern browser, such as Firefox, Opera, and Internet Explorer. However, we have tested the user interface currently only on Firefox and Opera under Linux.

Q: How does SWSE crawl the Semantic Web?

We used a version of MultiCrawler to crawl the Semantic Web. An ISWC publication about MultiCrawler is available online.

MultiCrawler transforms all crawled documents to RDF and creates an index over them. The focus is to find and transform Semantic Web documents, like RDF files and RSS feeds. To find URLs to Semantic Web files we extract URLs from the HTML Web, and follow rdfs:seeAlso links in RDF files.

Q: How many documents are indexed by SWSE?

Approximately data from about 400k sources, but we are constantly adding new sources.

Q: I thought you have 7 billion statements?

We stress-tested the system with 7 billion machine-generated statements. For the current demo, we use real-world data from the Semantic Web.

Q: How SWSE is implemented?

The components of SWSE are:

  1. Crawler (MultiCrawler) - Java
  2. Repository (YARS2) - Java
  3. Ranking (ReconRank) - Java
  4. Server-side User Interface - Java
  5. Client-side User Interface - XML/XSLT/CSS/JavaScript

Research

Q: Where can I find information on the research relating to SWSE?

More information on SWSE is available here.

Q: Are there any other systems like SWSE?

Yes. Other semantic web search engines include: