IDEAS home Printed from https://ideas.repec.org/a/spr/elmark/v32y2022i4d10.1007_s12525-022-00612-5.html
   My bibliography  Save this article

Global reconstruction of language models with linguistic rules – Explainable AI for online consumer reviews

Author

Listed:
  • Markus Binder

    (University of Regensburg, Germany, at the Faculty of Informatics and Data Science)

  • Bernd Heinrich

    (University of Regensburg, Germany, at the Faculty of Informatics and Data Science)

  • Marcus Hopf

    (University of Regensburg, Germany, at the Faculty of Informatics and Data Science)

  • Alexander Schiller

    (University of Regensburg, Germany, at the Faculty of Informatics and Data Science)

Abstract

Analyzing textual data by means of AI models has been recognized as highly relevant in information systems research and practice, since a vast amount of data on eCommerce platforms, review portals or social media is given in textual form. Here, language models such as BERT, which are deep learning AI models, constitute a breakthrough and achieve leading-edge results in many applications of text analytics such as sentiment analysis in online consumer reviews. However, these language models are “black boxes”: It is unclear how they arrive at their predictions. Yet, applications of language models, for instance, in eCommerce require checks and justifications by means of global reconstruction of their predictions, since the decisions based thereon can have large impacts or are even mandatory due to regulations such as the GDPR. To this end, we propose a novel XAI approach for global reconstructions of language model predictions for token-level classifications (e.g., aspect term detection) by means of linguistic rules based on NLP building blocks (e.g., part-of-speech). The approach is analyzed on different datasets of online consumer reviews and NLP tasks. Since our approach allows for different setups, we further are the first to analyze the trade-off between comprehensibility and fidelity of global reconstructions of language model predictions. With respect to this trade-off, we find that our approach indeed allows for balanced setups for global reconstructions of BERT’s predictions. Thus, our approach paves the way for a thorough understanding of language model predictions in text analytics. In practice, our approach can assist businesses in their decision-making and supports compliance with regulatory requirements.

Suggested Citation

  • Markus Binder & Bernd Heinrich & Marcus Hopf & Alexander Schiller, 2022. "Global reconstruction of language models with linguistic rules – Explainable AI for online consumer reviews," Electronic Markets, Springer;IIM University of St. Gallen, vol. 32(4), pages 2123-2138, December.
  • Handle: RePEc:spr:elmark:v:32:y:2022:i:4:d:10.1007_s12525-022-00612-5
    DOI: 10.1007/s12525-022-00612-5
    as

    Download full text from publisher

    File URL: http://link.springer.com/10.1007/s12525-022-00612-5
    File Function: Abstract
    Download Restriction: Access to the full text of the articles in this series is restricted.

    File URL: https://libkey.io/10.1007/s12525-022-00612-5?utm_source=ideas
    LibKey link: if access is restricted and if your library uses this service, LibKey will redirect you to where you can use your library subscription to access this item
    ---><---

    As the access to this document is restricted, you may want to search for a different version of it.

    References listed on IDEAS

    as
    1. James O’Donovan & Hannes F Wagner & Stefan Zeume, 2019. "The Value of Offshore Secrets: Evidence from the Panama Papers," The Review of Financial Studies, Society for Financial Studies, vol. 32(11), pages 4117-4155.
    2. Shrestha, Yash Raj & Krishna, Vaibhav & von Krogh, Georg, 2021. "Augmenting organizational decision-making with deep learning algorithms: Principles, promises, and challenges," Journal of Business Research, Elsevier, vol. 123(C), pages 588-603.
    3. Andreas J. Steur & Fabian Fritzsche & Mischa Seiter, 2022. "It’s all about the text: An experimental investigation of inconsistent reviews on restaurant booking platforms," Electronic Markets, Springer;IIM University of St. Gallen, vol. 32(3), pages 1187-1220, September.
    4. Chatterjee, Swagato & Goyal, Divesh & Prakash, Atul & Sharma, Jiwan, 2021. "Exploring healthcare/health-product ecommerce satisfaction: A text mining and machine learning application," Journal of Business Research, Elsevier, vol. 131(C), pages 815-825.
    5. Bernd Heinrich & Marcus Hopf & Daniel Lohninger & Alexander Schiller & Michael Szubartowicz, 2021. "Data quality in recommender systems: the impact of completeness of item content data on prediction accuracy of recommender systems," Electronic Markets, Springer;IIM University of St. Gallen, vol. 31(2), pages 389-409, June.
    Full references (including those not matched with items on IDEAS)

    Citations

    Citations are extracted by the CitEc Project, subscribe to its RSS feed for this item.
    as


    Cited by:

    1. Christian Meske & Babak Abedin & Mathias Klier & Fethi Rabhi, 2022. "Explainable and responsible artificial intelligence," Electronic Markets, Springer;IIM University of St. Gallen, vol. 32(4), pages 2103-2106, December.

    Most related items

    These are the items that most often cite the same works as this one and are cited by the same works as this one.
    1. Ashton, John & Burnett, Tim & Diaz-Rainey, Ivan & Ormosi, Peter, 2021. "Known unknowns: How much financial misconduct is detected and deterred?," Journal of International Financial Markets, Institutions and Money, Elsevier, vol. 74(C).
    2. Fangjun Wang & Shuolei Xu & Junqin Sun & Charles P. Cullinan, 2020. "Corporate Tax Avoidance: A Literature Review And Research Agenda," Journal of Economic Surveys, Wiley Blackwell, vol. 34(4), pages 793-811, September.
    3. Colonnelli, Emanuele & Lagaras, Spyridon & Ponticelli, Jacopo & Prem, Mounu & Tsoutsoura, Margarita, 2022. "Revealing corruption: Firm and worker level evidence from Brazil," Journal of Financial Economics, Elsevier, vol. 143(3), pages 1097-1119.
    4. Konda, Laura & Patel, Elena & Seegert, Nathan, 2022. "Tax enforcement and the intended and unintended consequences of information disclosure," Journal of Public Economics, Elsevier, vol. 212(C).
    5. Sharafutdinova,Gulnaz & Lokshin,Michael M., 2020. "Hide and Protect : A Role of Global Financial Secrecy in Shaping Domestic Institutions," Policy Research Working Paper Series 9348, The World Bank.
    6. Laila Ait Bihi Ouali, 2020. "Effects of signalling tax evasion on redistribution and voting preferences: Evidence from the Panama Papers," PLOS ONE, Public Library of Science, vol. 15(3), pages 1-22, March.
    7. Wen Zhang & Qiang Wang & Jian Li & Zhenzhong Ma & Gokul Bhandari & Rui Peng, 2023. "What makes deceptive online reviews? A linguistic analysis perspective," Palgrave Communications, Palgrave Macmillan, vol. 10(1), pages 1-14, December.
    8. Nilashi, Mehrbakhsh & Abumalloh, Rabab Ali & Samad, Sarminah & Alrizq, Mesfer & Alyami, Sultan & Abosaq, Hamad & Alghamdi, Abdullah & Akib, Noor Adelyna Mohammed, 2022. "Factors impacting customer purchase intention of smart home security systems: Social data analysis using machine learning techniques," Technology in Society, Elsevier, vol. 71(C).
    9. Ferris, Stephen P. & Hanousek, Jan & Tresl, Jiri, 2021. "Corporate profitability and the global persistence of corruption," Journal of Corporate Finance, Elsevier, vol. 66(C).
    10. Chenfeng Yan & Quan Chen & Xinyue Zhou & Xin Dai & Zhilin Yang, 2024. "When the Automated fire Backfires: The Adoption of Algorithm-based HR Decision-making Could Induce Consumer’s Unfavorable Ethicality Inferences of the Company," Journal of Business Ethics, Springer, vol. 190(4), pages 841-859, April.
    11. Dutt, Verena & Spengel, Christoph & Vay, Heiko, 2021. "Der EU-Vorschlag zum Country-by-Country Reporting im Internet: Kosten, Nutzen, Konsequenzen," Studien, Stiftung Familienunternehmen / Foundation for Family Businesses, number 250010.
    12. Verena K. Dutt & Christopher A. Ludwig & Katharina Nicolay & Heiko Vay & Johannes Voget, 2019. "Increasing tax transparency: investor reactions to the country-by-country reporting requirement for EU financial institutions," International Tax and Public Finance, Springer;International Institute of Public Finance, vol. 26(6), pages 1259-1290, December.
    13. Mahmoud Abdulhadi Alabdali & Sami A. Khan & Muhammad Zafar Yaqub & Mohammed Awad Alshahrani, 2024. "Harnessing the Power of Algorithmic Human Resource Management and Human Resource Strategic Decision-Making for Achieving Organizational Success: An Empirical Analysis," Sustainability, MDPI, vol. 16(11), pages 1-30, June.
    14. Juite Wang & Tzu-Yen Hsu, 2023. "Early discovery of emerging multi-technology convergence for analyzing technology opportunities from patent data: the case of smart health," Scientometrics, Springer;Akadémiai Kiadó, vol. 128(8), pages 4167-4196, August.
    15. Oliveira, Fabio & Kakabadse, Nada & Khan, Nadeem, 2022. "Board engagement with digital technologies: A resource dependence framework," Journal of Business Research, Elsevier, vol. 139(C), pages 804-818.
    16. repec:hal:wpspec:info:hdl:2441/5stdfj99hg96f8n537oi1mkfo9 is not listed on IDEAS
    17. repec:hal:spmain:info:hdl:2441/5stdfj99hg96f8n537oi1mkfo9 is not listed on IDEAS
    18. Agustín Bénétrix & Lorenz Emter & Martin Schmitz, 2024. "Automatic for the (tax) people: information sharing and cross-border investment in tax havens," Economic Policy, CEPR, CESifo, Sciences Po;CES;MSH, vol. 39(120), pages 853-895.
    19. Do, Quoc-Anh & Galbiati, Roberto & Marx, Benjamin & Ortiz Serrano, Miguel A., 2024. "J'Accuse! Antisemitism and financial markets in the time of the Dreyfus Affair," Journal of Financial Economics, Elsevier, vol. 154(C).
    20. Ruan, Yanya & Mezei, József, 2022. "When do AI chatbots lead to higher customer satisfaction than human frontline employees in online shopping assistance? Considering product attribute type," Journal of Retailing and Consumer Services, Elsevier, vol. 68(C).
    21. Fernando, Garcia Alvarado & Antoine, Mandel, 2022. "The network structure of global tax evasion evidence from the Panama papers," Journal of Economic Behavior & Organization, Elsevier, vol. 197(C), pages 660-684.
    22. François-Xavier de Vaujany & Stefan Haefliger & Paula Ungureanu, 2022. "From Collaborative Spaces to New Modes of Organizing: Society, Democracy and Commons on the Way to Novelty [Des espaces collaboratifs aux nouvelles formes d'organisation : société, démocratie et co," Post-Print hal-03827462, HAL.

    More about this item

    Keywords

    Explainable AI; Text analytics; Language models; BERT; Linguistic rules; Online consumer reviews;
    All these keywords.

    JEL classification:

    • C80 - Mathematical and Quantitative Methods - - Data Collection and Data Estimation Methodology; Computer Programs - - - General

    Statistics

    Access and download statistics

    Corrections

    All material on this site has been provided by the respective publishers and authors. You can help correct errors and omissions. When requesting a correction, please mention this item's handle: RePEc:spr:elmark:v:32:y:2022:i:4:d:10.1007_s12525-022-00612-5. See general information about how to correct material in RePEc.

    If you have authored this item and are not yet registered with RePEc, we encourage you to do it here. This allows to link your profile to this item. It also allows you to accept potential citations to this item that we are uncertain about.

    If CitEc recognized a bibliographic reference but did not link an item in RePEc to it, you can help with this form .

    If you know of missing items citing this one, you can help us creating those links by adding the relevant references in the same way as above, for each refering item. If you are a registered author of this item, you may also want to check the "citations" tab in your RePEc Author Service profile, as there may be some citations waiting for confirmation.

    For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: Sonal Shukla or Springer Nature Abstracting and Indexing (email available below). General contact details of provider: http://www.springer.com .

    Please note that corrections may take a couple of weeks to filter through the various RePEc services.

    IDEAS is a RePEc service. RePEc uses bibliographic data supplied by the respective publishers.