Publication: SARCASM DETECTION IN SPAM REVIEWS, A SYSTEMATIC LITERATURE REVIEW
| dc.contributor.advisor | Koç, Hasan | |
| dc.contributor.author | Capinig, Jaeden | |
| dc.contributor.department | Data Science and Business (BA) | |
| dc.contributor.other | Berlin International University of Applied Sciences | |
| dc.date.accessioned | 2025-12-10T07:53:59Z | |
| dc.date.available | 2025-12-10T07:53:59Z | |
| dc.date.issued | 2025 | |
| dc.description.abstract | Purpose – This systematic literature review investigates the current state of research on sarcasm detection in spam reviews, with the aim of informing businesses that utilise automated spam detection systems. Better understanding of this topic may help refine prediction accuracy and enable more reliable interpretation of customer feedback. Methodology – The review was conducted using a combined approach that integrates the principles outlined by Bandara et al. (2015) and the PRISMA framework, ensuring both methodological rigour and transparent reporting. Studies were selected based on clearly defined inclusion and exclusion criteria, and qualitative data was analysed using coding and thematic analysis techniques. Findings – The analysis revealed that contextual information is a critical feature in accurately detecting sarcasm, as sarcasm often relies on nuanced cues beyond surface-level sentiment. A variety of machine learning and deep learning models have been identified in the literature, with hybrid approaches (combining traditional and modern techniques) showing promise in improving detection accuracy. | |
| dc.description.degree | BA | |
| dc.identifier.uri | https://repository.berlin-international.de/handle/123456789/1222 | |
| dc.subject | sarcasm | |
| dc.subject | sarcasm detection | |
| dc.subject | spam | |
| dc.subject | spam detection | |
| dc.subject | online reviews | |
| dc.title | SARCASM DETECTION IN SPAM REVIEWS, A SYSTEMATIC LITERATURE REVIEW | |
| dc.type | Thesis | |
| dspace.entity.type | Publication | |
| relation.isAdvisorOfPublication | 92ccdeae-8cba-4697-892a-fb58ada8e886 | |
| relation.isAdvisorOfPublication.latestForDiscovery | 92ccdeae-8cba-4697-892a-fb58ada8e886 |