Inferring implications in semantic maps via the Apriori algorithm

Publication date: Available online 1 February 2020Source: LinguaAuthor(s): Ni Zeng, Hong XiaoAbstractSemantic maps are powerful tools for analyzing cross-language variations with implications between semantic functions to construct the relevant conceptual space. However, as existing semantic maps cannot illustrate the imbalance of implications between functions, a further discussion of inferring implications is highly demanded. The problem of inferring implications and the imbalance of implications between functions above is similar to the well-known problem of generating all significant association rules between items purchased by customers, and the Apriori algorithm offers an effective solution to relieve such issue. Here, alternative schematic diagrams based on the Apriori algorithm are employed to supplement semantic maps, which is justified by reproducing the same results as Cysouw obtained on person marking using his datasets. Furthermore, in our study, an implication in number from singular to plural is observed in person marking. We have solved the issue of imbalance and obtained credible implication rules between primitives in groups. We can mine practicable directed implication rules and reveal rules hard to notice before, such as active primitives between groups.

Source: Lingua - February 2, 2020 Category: Speech-Language Pathology Source Type: research