1 |
Methods and models in historical comparative research on signed languages ...
|
|
|
|
BASE
|
|
Show details
|
|
2 |
Methods and models in historical comparative research on signed languages
|
|
|
|
BASE
|
|
Show details
|
|
4 |
The origins of Russian-Tajik Sign Language : investigating the historical sources and transmission of a signed language in Tajikistan
|
|
|
|
BASE
|
|
Show details
|
|
5 |
An analysis and reconstruction of transitive nominalization in Ch’olan languages
|
|
|
|
BASE
|
|
Show details
|
|
6 |
Can the Comparative Method be used for signed language historical analyses? ...
|
|
|
|
BASE
|
|
Show details
|
|
7 |
Lexical conventionalization and the emergence of grammatical devices in a second generation homesign system in Peru
|
|
|
|
BASE
|
|
Show details
|
|
8 |
The phonetics, phonology, and morphology of Chajul Ixil (Mayan)
|
|
|
|
BASE
|
|
Show details
|
|
9 |
Computational models of changes in language use
|
|
|
|
Abstract:
There has been a large body of research on distributional models, which are computational models of word use. However, less research has explored diachronic distributional models, which predict changes in word use over time. In this dissertation, we expand upon this research in several ways. We propose several diachronic distributional models that, instead of having to break up the data into bins, represent time as a continuous variable. The first of these is a deep neural network that incorporates time via a feedforward network. The second is an RBF network that provides more interpretable information on how a word changes usage over time. We then propose several novel methods for evaluating diachronic distributional models. One is a pseudoword task where a model predicts synthetic changes in meaning. The other is a collection of tasks where a model predicts how a word changed senses over time. We then expand the scope of our analysis beyond diachronic distributional models. One direction we pursue is the development of a distributional model that incorporates regional information instead of temporal information. We then use this model to predict the linguistic regions of Texas. A second direction we pursue is the use of epidemiological models to explore sources of changes in word use ; Linguistics
|
|
Keyword:
Computational linguistics; Linguistics; NLP
|
|
URL: https://doi.org/10.26153/tsw/12537 https://hdl.handle.net/2152/85573
|
|
BASE
|
|
Hide details
|
|
11 |
Contact-induced grammaticalization as an impetus for arabic dialect development
|
|
|
|
BASE
|
|
Show details
|
|
12 |
Sculpting the narrative : the material practice of Epi-Olmec art and writing
|
|
|
|
BASE
|
|
Show details
|
|
15 |
Points of comparison : what indicating gestures tell us about the origins of signs in San Juan Quiahije Chatino sign language
|
|
|
|
BASE
|
|
Show details
|
|
17 |
An investigation of projection and temporal reference in Kaqchikel
|
|
|
|
BASE
|
|
Show details
|
|
18 |
"Making hands" : family sign languages in the San Juan Quiahije community
|
|
|
|
BASE
|
|
Show details
|
|
20 |
Language contact, inherited similarity and social difference : the story of linguistic interaction in the Maya Lowlands
|
|
|
|
MPI-SHH Linguistik
|
|
Show details
|
|
|
|