Question 1
Prior to the World Wide Web has been developed, information access was difficult and demanded expert knowledge on how to use these file transferring protocols, searching for information was a tedious work.
Document centered processing has not changed with the WWW, and browsers just arose with the WWW.
Question 2
Keyword-based search leads to irrelevant and incomplete results, while search queries are simple and the amount of indexed documents can be enormously big.
Question 3
Web 2.0 changed the way users interacted with the Web. Interactive contents are fancy stuff that existed long ago, and correct interpretation of data by computers is a vision that is pursued by researchers, e.g. with Semantic Web technology.
Question 4
Asking a search engine for "largest eu city" might bring you to the Wikipedia page of Largest cities of the European Union. There you see the largest city is London with 8,308,369 inhabitants. The correct answer is only one click away, you find the population density within the infobox of the article on London is 5,285/sq km, approx. 5,300/sq km is the correct answer.
If you found the Largest cities in Europe list on Wikipedia, you'd have needed to skip Instanbul and Moscow, since both cities are not in the European Union.
You easily see how this question could easily be answered with querying an appropriate database.
Regrading Since some web resources (e.g. the Dutch Wikipedia and the London Council website) reported the population density to be 4.761 inw/sq km or 12,331/sq mi, the answer "approx. 4,780/sq km (12,380/sq mi)" should be correct as well. Anybody, who submitted this answer will get the points. Therefore, we will change these answers to the intended one. Don't be surprised, when your answer changes to the intended one.
Question 5
Please do not hit on us, for putting this brain-teaser in the weekly assignment. We knew it is a hard query for keyword-based search, and we hope some challenged it and it has been an interesting challenge to all of you.
What we are searching for is obviously a place or location. It is extra-terrestrical, since not on planet Earth. It is on some kind of globe, since located in a hemisphere.
If you ask a search engine for "Lod", you will probably get links to documents about Lod - a town in Israel, or LOD as an abbreviation for "Level Of Detail", "Linked Open Data", "Lord Of Destruction", "Length Of Day", and many more. You might also have found the disambiguation page of Lod, that shows "all" concepts with that particular name (Well, it doesn't!).
It can't be a town in Israel, since that one is on Earth. It also can't be one of the abbreviations of abstract concepts, since that are not locatable.
Googling "lod not on earth" even returns spell-corrected results, such as "How old is the earth?".
So what do do? Use your background knowledge!
If it is not on Earth, it might be on another planet. So, we could try queries like "lod mercury", "lod venus", "lod mars", ... and tadaa, you will find the correct entity: Lod is a crater on planet Mars, and it is located 21.2°N 31.6°W, i.e. on the Northern hemisphere.
Further ways:
- If you ask for "lod northern hemisphere", you'll find the Wikipedia article about Geology of Mars by incident, there you can find a link to the crater.
- Looking at previous versions of the Wikipedia Lod (disambiguation) page, someone already added the mars crater and undone this change :(
You see how challenging it can be to find an entity by it's name. A knowledge base containing names (possibly also alternative names) and further information on any entities might help. Try this DBpedia query.
Question 6
The Semantic Web Stack provides an architecture for the Semantic Web and a bunch of technologies standardized by the W3C. RDF allows to create statements in form of subject-predicate-object triples.
It is no substitute to the classical hypertext web and SQL is not part of the Semantic Web Stack.
To build Semantic Web applications it is sufficient to select the technologies according to the needs and abstraction level of the application.
Question 7
The role of ontologies in the Semantic Web is to express the knowledge of a particular domain, i.e. to describe classes, properties, and their relationships. This descriptions can be applied to multiple entities and allow to deduce new knowledge by logical inference.