Data Ethics Club meeting 28-02-24, 1pm UK time#

Meeting info#

Description#

You’re welcome to join us for our next Data Ethics Club meeting on 28th Feb 2024 at 1pm UK time. You don’t need to register, just pop in. This time we’re going to watch/read Google Search Really Has Gotten Worse, Researchers Find by Jason Koebler, which discusses a research paper that finds a lot of spam making its way into search results of product reviews.

Thank you to Euan Bennet for suggesting this week’s content.

You can read this article on 404 media by subscribing for free or read the original research paper that it was based on by Bevendorff et al which will be presented at ECIR in March.

Summary#

The researchers analysed search results for 7392 product reviews on three different search engines: Google, Bing and DuckDuckGo. They monitored results for these product review searches generating and archiving results every two weeks for one year. They find several issues among the higher-ranked pages:

  • they often contain more affiliate marketing than the average product review

  • they often show signs of lower text quality e.g. by using repetitive vocabulary

  • spam sites and review farms (i.e. seemingly generated listicles without any editorial content) keep finding ways to creep to the top result

These websites leverage known SEO (search engine optimisation) features such as length of the content and URL depth, i.e. how many files follow after the domain. All of this will probably get worse as AI-generated spam will swamp the internet, which is a trend the researchers are starting to observe. They can also see search engines reacting by kicking spam sites of high ranks with every update, but some spam sites manage to make a comeback over their year long analysis period.

In a reply, Google pointed out that the study only focuses on the narrow application area of product review content and that they performed better than other search engines.

Discussion points#

There will be time to talk about whatever we like, relating to the paper, but here are some specific questions to think about while you’re reading.

  • Does this research reflect your own experience of how search engines have evolved? What about for searches outside of product reviews?

  • The 404 article also refers to other investigations of search engine optimisation such as one from the Verge that “blamed it for ruining the internet”. Do you agree? Who or what do you think is responsible for the decreasing quality of search results?

  • Can you imagine better, more equitable or more useful features that could guide search engine optimisation in making the web searchable?