Towards Explicit Semantic Features using Thresholded Independent Component Analysis
نویسندگان
چکیده
Latent semantic analysis (LSA) can be used to create an implicit semantic vectorial representation for words. Independent component analysis (ICA) can be derived as an extension to LSA that rotates the latent semantic space so that it becomes explicit, that is, the features correspond more with those resulting from human cognitive activity. This enables nonlinear filtering of the features, such as hard thresholding that creates a sparse word representation where only a subset of the features is required to represent each word successfully. We demonstrate this with semantic multiple choice vocabulary tests. The experiments are conducted in English, Finnish and Swedish.
منابع مشابه
Explicit vs. Contrastive-based Instruction of Formulaic Expressions in Developing EFL Learners’ Reading Ability
As an integrative component of textual structure, formulaic expressions (FEs) play a key role in communicating the message and comprehending the text. Furthermore, interlingually contrastive features of FEs add to their both significance and complexity of their instruction. Given these facts, this study was an attempt to explore a sound mechanism on how to teach FEs; whether an explicit or CA-...
متن کاملEmergence of Linguistic Features: Independent Component Analysis of Contexts
We show that independent component analysis (ICA) (Hyvärinen et al. 2001) applied on word context data gives distinct features that reflect syntactic and semantic categories. The analysis gives features or categories that are both explicit and can easily be interpreted by humans. This result can be obtained without any human supervision or tagged corpora that would have some predetermined morph...
متن کاملInvestigating the mechanism of the void's physical-semantic effect on social interactions
The depth of the void concept has extended the range of its effects from philosophy to various sciences and even types of art. In architecture, due to the importance of the spacing effect and architectural components on behavior, void finds a different role that seems to be less addressed in contemporary architecture. If void, regardless of its hidden meaning, is referred to as "empty space," a...
متن کاملSemantic analysis in word vector spaces with ICA and feature selection
In this article, we test a word vector space model using direct evaluation methods. We show that independent component analysis is able to automatically produce meaningful components that correspond to semantic category labels. We also study the amount of features needed to represent a category using feature selection with syntactic and semantic category test sets.
متن کاملThresholded Multivariate Principal Component Analysis for Phase I Multichannel Profile Monitoring
Monitoring multichannel profiles has important applications in manufacturing systems improvement, but it is non-trivial to develop efficient statistical methods due to the facts that profiles are high-dimensional functional data with intrinsic innerand inter-channel correlations, and that the change might only affect a few unknown features of multichannel profiles. To tackle these challenges, w...
متن کامل