idw – Informationsdienst Wissenschaft

Nachrichten, Termine, Experten

Grafik: idw-Logo
Medienpartner:
Wissenschaftsjahr


Share on: 
09/08/2017 10:03

First results of the Bundestag election study: The Google search engine hardly personalises

Melanie Löw Presse- und Öffentlichkeitsarbeit
Technische Universität Kaiserslautern

    Which results are shown by the Google search engine when asked about politicians, parties and the Bundestag election? This is a question currently being asked by the team of computer science professor Dr Katharina Anna Zweig of the Technical University Kaiserslautern, in collaboration with the non-governmental organisation (NGO) AlgorithmWatch. So far, the project has analysed over three million datasets from more than 4,000 individuals. First results show that the results in the hit lists are hardly personalised at all. When it comes to searching for politicians, the results lists were almost identical, and the majority of results for the parties also overlap.

    The project is being funded by six federal state media authorities.

    For the Bundestag election campaign, the Zweig’s team is investigating for the first time for whom which news or results are displayed by the Google search engine provider. As such, it is working closely with AlgorithmWatch, and put out an appeal for people to donate their data several weeks ago. More than 4,000 people participated in the study. Using a software expansion, which was made available among others by the news portal Spiegel Online, the media partner of the project, the researchers were able to send search queries automatically from the participants’ computers up to six times a day. As a result, between 300 and 600 search results on 16 search terms – a total of three million datasets – were generated every day.

    Now the first analyses are available. “For the search results, there is only limited space for personalisation,” says Professor Zweig from the working group “Algorithm Accountability Lab” at the University of Kaiserslautern. For politicians in particular, the Google hit lists show a high level of similarity. “On average, there are seven or eight identical results for nine hits,” says Zweig. For the parties, the similarity is slightly lower with five to six identical results. “But this is due primarily to regional aspects. Here too, only one to two links remain on average for personalisation,” the computer scientist continues.

    It also makes no difference whether someone is logged in to their Google account or uses the search engine anonymously. The hit list is almost identical for both groups.

    However, it also appears that the parties vary in their success in placing their own websites and social media channels onto the first page of the Google hit list: The most successful is the party Bündnis 90/Die Grüne. “This party’s own content is represented with an almost constant rate of 70 percent,” says Zweig. They are followed by Die Linke, FDP and CDU with 60 percent, CSU and SPD with 40 percent, and the AfD with 30 percent.

    It will be seen over the coming weeks whether the results change before the Bundestag election. “The project shows that society can jointly investigate a major algorithm, which can potentially influence the public and opinion formation, in terms of relevant phenomena, without requiring access to the underlying code,” says Zweig with regard to the study.

    The work „#Datenspende: Google und die Bundestagswahl 2017“ is being carried out by the Kaiserslautern research team in collaboration with AlgorithmWatch, whose co-founder is Zweig. Furthermore, the federal state media authorities of Bavaria (BLM), Berlin-Brandenburg (mabb), Hessen (LPR Hessen), Rhineland-Palatinate (LMK), Saarland (LMS) and Saxony (SLM) are involved in the study. Spiegel Online is the acting media partner.

    The detailed results of the study “Filterblase geplatzt? Kaum Raum für Personalisierung bei Google-Suchen zur Bundestagswahl 2017” (“Filter bubble burst? Not much space for personalisation in Google searches on the Bundestag election 2017”) can be found at: algorithmwatch.org
    More information on the project at datenspende.algorithmwatch.org

    Please direct enquiries to:
    Professor Dr Katharina Anna Zweig
    E-Mail: zweig(at)cs.uni-kl.de
    Phone: 0631 205-3346

    Tobias Krafft
    E-Mail: krafft(at)cs.uni-kl.de
    Phone: 0631 205-4070


    Criteria of this press release:
    Business and commerce, Journalists
    Information technology, Media and communication sciences, Politics
    transregional, national
    Research projects, Research results
    English


    Prof. Katharina Anna Zweig


    For download

    x

    Help

    Search / advanced search of the idw archives
    Combination of search terms

    You can combine search terms with and, or and/or not, e.g. Philo not logy.

    Brackets

    You can use brackets to separate combinations from each other, e.g. (Philo not logy) or (Psycho and logy).

    Phrases

    Coherent groups of words will be located as complete phrases if you put them into quotation marks, e.g. “Federal Republic of Germany”.

    Selection criteria

    You can also use the advanced search without entering search terms. It will then follow the criteria you have selected (e.g. country or subject area).

    If you have not selected any criteria in a given category, the entire category will be searched (e.g. all subject areas or all countries).