No Keyword is an Island: In search of covert associations
This paper describes how corpus-assisted discourse analysis based on keyword (KW) identification and interpretation can benefit from employing Market basket analysis (MBA) after KW extraction. MBA is a data mining technique used originally in marketing that can reveal consistent associations between items in a shopping cart, but also between keywords in a corpus of many texts. By identifying recurring associations between KWs we can compensate for the lack of wider context which is a major issue impeding the interpretation of isolated KWs (esp. when analyzing large data). To showcase the advantages of MBA in "re-contextualizing" keywords within the discourse, a pilot study on the topic of migration was conducted contrasting anti-system and center-right Czech internet media. was conducted. The results show that MBA is useful in identifying the dominant strategy of anti-system news portals: to weave in a confounding ideological undercurrent and connect the concept of migrants to a multitude of other topics (i.e., flooding the discourse).
Code (0)
등록된 구현이 없습니다.
Tasks
MarketingSimilar Papers 제목 키워드 기반
Unveiling Covert Toxicity in Multimodal Data via Toxicity Association Graphs: A Graph-Based Metric and Interpretable Detection Framework
Detecting toxicity in multimodal data remains a significant challenge, as harmful meanings often lurk beneath seemingly benign individual modalities: only emerging when modalities are combined and semantic associations a…
Higher serum 25(OH)D concentration is associated with lower risk of metabolic syndrome among Aboriginal and Torres Strait Islander peoples in Australia
Although previous observational studies have shown associations between serum 25-hydroxyvitamin D (25(OH)D) concentration and metabolic syndrome, this association has not yet been investigated among Aboriginal and Torres…
Cordyceps: Covert Control Attacks on LLMs via Data Poisoning
Large language models (LLMs) are often fine-tuned on uncurated text datasets that adversaries can poison. Existing poisoning attacks primarily rely on fixed trigger phrases that defenses such as outlier detection, clean-…
Outlier DetectionAutoKG: Efficient Automated Knowledge Graph Generation for Language Models
Traditional methods of linking large language models (LLMs) to knowledge bases via the semantic similarity search often fall short of capturing complex relational dynamics. To address these limitations, we introduce Auto…
Graph GenerationRetrievalSemantic SimilaritySemantic Textual SimilarityMining User Opinions in Mobile App Reviews: A Keyword-based Approach
User reviews of mobile apps often contain complaints or suggestions which are valuable for app developers to improve user experience and satisfaction. However, due to the large volume and noisy-nature of those reviews, m…