Results

Found 2 articles

Clear filters

Global reconstruction of language models with linguistic rules – Explainable AI for online consumer reviews
PDF
2022 | Electronic Markets | Citations: 0
Authors: Binder, Markus; Heinrich, Bernd; Hopf, Marcus; Schiller, Alexander

Abstract: Analyzing textual data by means of AI models has been recognized as highly relev ... Expand

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. Collapse

Semantic filters: Stanza

Topics: explainable artificial intelligence laptop computer experience good electronic market artificial intelligence

Methods: BERT language model natural language processing machine learning product review
Knowledge is Power: Provide Your IT-Support with Domain-Specific High-Quality Solution Material
PDF
2021 | DESRIST | Citations: 0
Authors: Schmidt, Simon L.; Li, Mahei Manhai; Weigel, Sascha; Peters, Christoph

Abstract: As more and more business processes are based on IT services the high availabili ... Expand

Abstract: As more and more business processes are based on IT services the high availability of these processes is dependent on the IT-Support. Thus, making the IT-Support a critical success factor of companies. This paper presents how this department can be supported by providing the staff with domain-specific and high-quality solution material to help employees faster when errors occur. The solution material is based on previously solved tickets because these contain precise domain-specific solutions narrowed down to e.g., specific versions and configurations of hard-/software used in the company. To retrieve the solution material ontologies are used that contain the domain-specific vocabulary needed. Because not all previously solved tickets contain high-quality solution material that helps the staff to fix issues the de-signed IT-Support system separates low- from high-quality solution material. This paper presents (a) theory- and practical-motivated design requirements that describe the need for automatically retrieved solution material, (b) develops two major design principles to retrieve domain-specific and high-quality solution material, and (c) evaluates the instantiations of them as a prototype with organic real-world data. The results show that previously solved tickets of a company can be pre-processed and retrieved to IT-Support staff based on their current queries. Collapse

Semantic filters: Stanza

Topics: knowledge representation system support Python web application usability

Methods: ontological modelling computational algorithm design artifact natural language processing logistic regression

Page 1 of 1

AIS11

Journals

Conferences

Semantic Filter

Found 2 articles

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

Knowledge is Power: Provide Your IT-Support with Domain-Specific High-Quality Solution Material