The World Wide Web & Search Engines

Bottleneck 2: The World Wide Web & Search Engines

==============================================
Bottleneck 2: The World Wide Web & Search Engines
The World Wide Web (WWW) is an unstructured, de-centralized and
ever changing ecosystem. Google tries to make sense of it with different
types of software:
- Crawling software: bots are fast and greedy but blind to whatever
is not text. Search engines try to cope with users creating
incomplete, biased or inadequate connections between pieces of
information. Google in particular is trying to cope with the surge of
social networks.
The quality of links as indicators of relevancy or popularity is also
evolving: it is more convenient to link to content via social shares
than with traditional hyperlinks on web pages.
- Indexing software: the main current challenges are identifying
unique content and attributing authorship. Valuable content is
often very duplicated across the WWW.
- Ranking software: Google claims to fight web spam with training
data built for machine learning algorithms allegedly used for userspecific and session-specific rankings.
++++++++++++++++++++++++++++++++++++++++++++++++++++++