What has the wife of a former German president and Google got in common?
Well, probably more than you might at first suspect. The search engine giant and the former first lady may well walk different paths, but the one thing they share is that they are both inextricably bound by the search engine results. Bettina Wulff is suing Google to stop rumours about her private life after false allegations appeared about her somewhat dubious past life on Google’s search engine results pages.
When the name Bettina Wolff is typed into Google’s search engine, suggested search terms include the words “prostitute” and “red light district”. Why someone would want to blacken Mrs Wolff’s name is unclear, though many German newspapers have reported that this online misinformation campaign was started in order to disrupt her husband Christian Wulff’s political career. Whatever the reason, the upshot is that both she and her partner have been tarnished by the allegations and they hold the search giant accountable.
Google, naturally denies any wrongdoing or culpability, and maintains that any auto-generated text results shown on its pages are merely reflective of what users are searching for.
So how has the problem arisen?
Users asked for more relevant and targeted search engine results, and Google promptly delivered.
So how has woman in Germany become a casualty?
Google’s algorithm is constantly evolving as it seeks to find new ways to make its results pages more relevant, whilst screening out as much spam as it can identify. This evolution led to the introduction of the concept of semantic search. Semantic search uses semantics, or the science of meaning in language, to produce highly relevant search results, rather than simply using ranking algorithms such as Google’s PageRank to predict relevancy. The goal is to deliver the information queried by a user, rather than have a user sort through a list of loosely related keyword results. Semantic search seeks to improve search accuracy by understanding searcher intent and the contextual meaning of terms as they appear in the searchable data-space, whether on the Web or within a closed system, to generate more apposite results.
Semantic search systems consider various points of reference including context of search, location, intent, the variation of words, synonyms, generalised and specialised queries, concept matching and natural language queries to provide relevant search results. Basic search built on keywords helps users navigate to particular pages when they’re searching for products or services. Semantic search will help users to research topics that they have little prior knowledge of: users provide the search engine with a phrase about a particular subject/object, and the engine will come up with a list of documents which the user can then use to gain knowledge about the subject they’re researching.
The only problem appears to be that sometimes you have to be careful what you wish for, because when you get it there may well be a sting in the tail. Search engine algorithm results can be manipulated; hence Google’s crack down on black hatpractices. However it seems that predictive searching through semantics can also be twisted to produce distorted results just as easily. If a person wants to deliberately cause harm or blacken their reputation, then all that needs to be done is start a malicious rumour or two and eventually the search engines will inadvertently pick up on this and feed the flame.
So what has Google had to say about the matter?
Well, Head of PR for Google Germany, Kay Oberbeck, told the BBC that the site’s search terms were:
“algorithmically generated and include the popularity of the entered search terms. All terms that appear have been previously entered by Google users,” he added in a statement.
He was also keen to point out that search rivals Bing also generated the same type of search results. The same text is generated in rival search engine Bing.com.
So is this case just a one off? Is it just an unfortunate and unprecedented mistake? Google would no doubt argue that it is, but history wouldn’t necessarily back it up. It’s happened before. In March 2012 Google was ordered to disable the autocomplete function relating to search results for an unnamed man in Japan, who said his name was being associated with crimes he had not committed.