Background: Abstract sentence classification modelling is a natural language processing approach to the classification of sentences from academic literature abstracts as indicative of a study’s key literature characteristics commonly used as structured abstract headings, such as background, aims, method and results. Much work has been done to advance abstract sentence classification modelling capability in the computer science domain, however, there has been no examination on the acceptance of this technology by researchers in the context of conducting academic literature discovery. Aims: This study aims to explore acceptance of abstract sentence classification modelling capability by information systems researchers using the technology acceptance model theoretical framework, which was extended to include the indicators originating from the information retrieval and perceived web quality research domains. Method: A prototype system deploying abstract sentence classification modelling capability into a pseudo academic literature index interface was developed. A conceptual model grounded in the technology acceptance model was then produced after consideration of existing relevant literature, before a survey instrument was prepared using measurement items identified and extended from existing literature. Information systems researchers who had published in leading information systems conference proceedings between 2019 and 2021 were invited to participate in the survey, with their survey responses used to evaluate the measurement and structural models using partial least squares structural equational modelling. Results: Results indicated that the capability of the prototype abstract sentence classification modelling system to perform sentence classification, the appearance of the prototype system and supplementary system features all positively and significantly effected perceived usefulness and ease of using abstract sentence classification modelling. Perceived usefulness and ease of using abstract sentence classification modelling had a positive and significant effect on participant attitude towards abstract sentence classification modelling and intention to use abstract sentence classification modelling. Participant attitudes towards the concept of abstract sentence classification modelling also had a positive and significant effect on intention to use abstract sentence classification modelling.
Stead, Connor; Smith, Stephen; Busch, Peter; and Vatanasakdakul, Savanid, "Abstract Sentence Classification Acceptance" (2023). ACIS 2023 Proceedings. 120.