Trkic G00gle: Why and How Users Game Translation Algorithms

Individuals interact with algorithms in various ways. Users even game and circumvent algorithms so as to achieve favorable outcomes. This study aims to come to an understanding of how various stakeholders interact with each other in tricking algorithms, with a focus towards online review communities. We employed a mixed-method approach in order to explore how and why users write machine non-translatable reviews as well as how those encrypted messages are perceived by those receiving them. We found that users are able to find tactics to trick the algorithms in order to avoid censoring, to mitigate interpersonal burden, to protect privacy, and to provide authentic information for enabling the formation of informative review communities. They apply several linguistic and social strategies in this regard. Furthermore, users perceive encrypted messages as both more trustworthy and authentic. Based on these findings, we discuss implications for online review community and content moderation algorithms.

[1]  D. Simon,et al.  Lost in translation: A framework for analysing complexity of co-production settings in relation to epistemic communities, linguistic diversities and culture , 2019, Futures.

[2]  River-Crabbed Shitizens and Missing Knives: A Sociolinguistic Analysis of Trends in Chinese Language Use Online as a Result of Censorship , 2015 .

[3]  Jo Mackiewicz,et al.  Product Review Users' Perceptions of Review Quality: The Role of Credibility, Informativeness, and Readability , 2014, IEEE Transactions on Professional Communication.

[4]  Sophie Bishop,et al.  Managing visibility on YouTube through algorithmic gossip , 2019, New Media Soc..

[5]  Tarleton Gillespie,et al.  Algorithmically recognizable: Santorum’s Google problem, and Google’s Santorum problem , 2017, The Social Power of Algorithms.

[6]  J. Chiou,et al.  How Much can you Trust Online Information? Cues for Perceived Trustworthiness of Consumer-generated Online Information , 2011 .

[7]  Matthew Hutson.  Hackers easily fool artificial intelligences. , 2018, Science.

[8]  Susan R. Fussell,et al.  Beyond Translation: Design and Evaluation of an Emotional and Contextual Knowledge Interface for Foreign Language Social Media Posts , 2018, CHI.

[9]  Daniel J. Liebling,et al.  Unmet Needs and Opportunities for Mobile Translation AI , 2020, CHI.

[10]  Eric Gilbert,et al.  Algorithmically Bypassing Censorship on Sina Weibo with Nondeterministic Homophone Substitutions , 2015, ICWSM.

[11]  Kenneth I. Forster,et al.  Masked form priming with extreme transposition , 2008 .

[12]  Airi Lampinen,et al.  Gaming Algorithmic Hate-Speech Detection: Stakes, Parties, and Moves , 2020 .

[13]  Brent Waters,et al.  Witness encryption and its applications , 2013, STOC '13.

[14]  Andreas Holzinger,et al.  Interactive machine learning: experimental evidence for the human in the algorithmic loop , 2018, Applied Intelligence.

[15]  Karrie Karahalios,et al.  First I "like" it, then I hide it: Folk Theories of Social Feeds , 2016, CHI.

[16]  David G. Rand,et al.  Will the Crowd Game the Algorithm?: Using Layperson Judgments to Combat Misinformation on Social Media by Downranking Distrusted Sources , 2020, CHI.

[17]  Taewook Kim.  Love in Lyrics: An Exploration of Supporting Textual Manifestation of Affection in Social Messaging , 2019 .

[18]  Victoria D. Bush,et al.  What We Know and Don't Know about Online Word-of-Mouth: A Review and Synthesis of the Literature , 2014 .

[19]  Shagun Jhaver,et al.  Algorithmic Anxiety and Coping Strategies of Airbnb Hosts , 2018, CHI.

[20]  K. Forster,et al.  Masked priming with graphemically related forms: Repetition or partial activation? , 1987 .

[21]  Mohammad Hossein Jarrahi,et al.  Gig Platforms, Tensions, Alliances and Ecosystems , 2019, Proc. ACM Hum. Comput. Interact..

[22]  Desiderio Gutiérrez-Taño,et al.  What drives guests to misreport their experiences on Airbnb? A structural equation modelling approach , 2020, Current Issues in Tourism.

[23]  J. Trefil,et al.  The new dictionary of cultural literacy , 2002 .

[24]  Matthew L. Jensen,et al.  Credibility of Anonymous Online Product Reviews: A Language Expectancy Perspective , 2013, J. Manag. Inf. Syst..

[25]  Hao-Chuan Wang,et al.  Task Rebalancing: Improving Multilingual Communication with Native Speakers-Generated Highlights on Automated Transcripts , 2017, CSCW.

[26]  Brent J. Hecht,et al.  3 Stars on Yelp, 4 Stars on Google Maps , 2021, Proc. ACM Hum. Comput. Interact..

[27]  V. Braun,et al.  Using thematic analysis in psychology , 2006 .

[28]  S. Viswanathan,et al.  Who Forgoes Screening in Online Markets and Why? Evidence from Airbnb , 2019, MIS Q..

[29]  B. Latour.  On actor-network theory : A few clarifications , 1996 .

[30]  Ysabel Gerrard,et al.  Behind the Screen: Content Moderation in the Shadows of Social Media , 2019, New Media Soc..

[31]  Jack Halpern.  Lexicon-based Orthographic Disambiguation in CJK Intelligent Information Retrieval , 2002, ALR@COLING.

[32]  Taeyoung Lee,et al.  Us vs. Them: Understanding Artificial Intelligence Technophobia over the Google DeepMind Challenge Match , 2017, CHI.

[33]  Rebecca Gray,et al.  Understanding User Beliefs About Algorithmic Curation in the Facebook News Feed , 2015, CHI.

[34]  Bin Xu,et al.  How beliefs about the presence of machine translation impact multilingual collaborations , 2014, CSCW.

[35]  Jürgen Ziegler,et al.  Let Me Explain: Impact of Personal and Impersonal Explanations on Trust in Recommender Systems , 2019, CHI.

[36]  Sara B. Kiesler,et al.  Why do people seek anonymity on the internet?: informing policy and design , 2013, CHI.

[37]  J. Dijck.  The Culture of Connectivity: A Critical History of Social Media , 2013 .

[38]  Malte Ziewitz,et al.  Rethinking gaming: The ethical work of optimization in web search engines , 2019, Social studies of science.

[39]  Mor Naaman,et al.  AI-Mediated Communication: Definition, Research Agenda, and Ethical Considerations , 2020, J. Comput. Mediat. Commun..

[40]  Linda Bol,et al.  A Comparison of Anonymous versus Identifiable e-Peer Review on College Student Writing Performance and the Extent of Critical Feedback. , 2007 .

[41]  Neha Kumar,et al.  The mobile media actor-network in urban India , 2013, CHI.

[42]  Saad A. Alhoqail,et al.  How Online Product Reviews Affect Retail Sales: A Meta-analysis , 2014 .

[43]  Eun Ju Lee,et al.  Is the message the medium? How politicians' Twitter blunders affect perceived authenticity of Twitter communication , 2020, Comput. Hum. Behav..

[44]  Kyung Hyan Yoo,et al.  What Motivates Consumers to Write Online Travel Reviews? , 2008, J. Inf. Technol. Tour..

[45]  Blaine Nelson,et al.  Adversarial machine learning , 2019, AISec '11.

[46]  Raffaele Filieri.  What makes an online consumer review trustworthy , 2016 .

[47]  Heinrich Hußmann,et al.  When people and algorithms meet: user-reported problems in intelligent everyday applications , 2019, IUI.

[48]  Alys Young,et al.  Qualitative Research and Translation Dilemmas , 2004 .

[49]  Xiao Ma,et al.  When Do People Trust Their Social Groups? , 2019, CHI.

[50]  Ying Liu,et al.  The Value of Multi-Dimensional Rating Systems: Evidence from a Natural Experiment and Randomized Experiments , 2017, Manag. Sci..

[51]  Georgios Zervas,et al.  Fake It Till You Make It: Reputation, Competition, and Yelp Review Fraud , 2015, Manag. Sci..

[52]  Won Ik Cho,et al.  BEEP! Korean Corpus of Online News Comments for Toxic Speech Detection , 2020, SOCIALNLP.

[53]  Xiao Ma,et al.  AI-Mediated Communication: How the Perception that Profile Text was Written by AI Affects Trustworthiness , 2019, CHI.

[54]  Camilla Vásquez,et al.  If nearly all Airbnb reviews are positive, does that make them meaningless? , 2018 .

[55]  Ting-Hao Huang,et al.  Evorus: A Crowd-powered Conversational Assistant Built to Automate Itself Over Time , 2018, CHI.

[56]  N. Kerr,et al.  Ringelmann Revisited , 1981 .

[57]  Xiao Ma,et al.  Self-Disclosure and Perceived Trustworthiness of Airbnb Host Profiles , 2017, CSCW.

[58]  S. Lupker,et al.  Can CANISO activate CASINO? Transposed-letter similarity effects with nonadjacent letter positions ☆ , 2004 .

[59]  Kevin Duh,et al.  Robsut Wrod Reocginiton via Semi-Character Recurrent Neural Network , 2016, AAAI.

[60]  Jeung‐Ryeul Cho,et al.  Orthographic and Phonological Activation in the Semantic Processing of Korean Hanja and Hangul. , 1999 .

[61]  Eric Gilbert,et al.  User Attitudes towards Algorithmic Opacity and Transparency in Online Reviewing Platforms , 2019, CHI.

[62]  Karrie Karahalios,et al.  "Be Careful; Things Can Be Worse than They Appear": Understanding Biased Algorithms and Users' Behavior Around Them in Rating Platforms , 2017, ICWSM.

[63]  Susan R. Fussell,et al.  Same translation but different experience: the effects of highlighting on machine-translated conversations , 2013, CHI.

[64]  Michael Luca.  Reviews, Reputation, and Revenue: The Case of Yelp.Com , 2016 .

[65]  Joonhwan Lee,et al.  Bot in the Bunch: Facilitating Group Chat Discussion by Improving Efficiency and Participation with a Chatbot , 2020, CHI.

[66]  Benjamin L. Edelman,et al.  The market design and policy of online review platforms , 2017 .

[67]  Kelley Cotter,et al.  Playing the visibility game: How digital influencers and algorithms negotiate influence on Instagram , 2018, New Media Soc..

[68]  K. Blashki,et al.  Game geek's goss: linguistic creativity in young males within an online university forum (94/\/\3 933k’5 9055oneone) , 2005 .

[69]  Michael Luca,et al.  Designing Online Marketplaces: Trust and Reputation Mechanisms , 2016, Innovation Policy and the Economy.

[70]  N. Lefkowitz.  Verlan: Talking Backwards in French , 1989 .

[71]  Andreas Butz,et al.  How to Trick AI: Users' Strategies for Protecting Themselves from Automatic Personality Assessment , 2020, CHI.

[72]  Barry A. T. Brown,et al.  Market Design for HCI: Successes and Failures of Peer-to-Peer Exchange Platforms , 2017, CHI.