Showing papers in "IEICE technical report. Speech in 2015"

Word translation entropy : evidence of early target language activation during reading for translation (Thought and Language)

Abstract: This study reports on an investigation into the relationship between the number of translation alternatives for a single word and eye movements on the source text. In addition, the effect of word order differences between source and target text on eye movements on the source text is studied. In particular, the current study investigates the effect of these variables on early and late eye movement measures. Early eye movement measures are indicative of processes that are more automatic while late measures are more indicative of conscious processing. Most studies that found evidence of target language activation during source text reading in translation, i.e. co-activation of the two linguistic systems, employed late eye movement measures or reaction times. The current study therefore aims to investigate if and to what extent earlier eye movement measures in reading for translation show evidence of co-activation. Results show that the number of translation alternatives for a single word and differences between source and target text in terms of word order have an effect on very early and late eye movement measures. Results are interpreted in terms of semantic and structural cross-linguistic priming: items which have a similar word order in source and target texts are likely to have similar syntactic structures. These items are therefore more likely to prime structurally. Source items which have few translation alternatives are more likely to share a semantic representation and are hence more likely to prime semantically than items with more translation alternatives. Findings support the literal translation hypothesis.

Computational Complexity Studies of Synchronous Boolean Finite Dynamical Systems (コンピュテーション)

Abstract: The finite dynamical system is a system consisting of some finite number of objects that take upon a value from some domain as a state, in which after initialization the states of the objects are updated based upon the states of the other objects and themselves according to a certain update schedule. This paper studies the subclass of finite dynamical systems the synchronous boolean finite dynamical system (synchronous BFDS, for short), where the states are boolean and the state update takes place in discrete time and at the same on all objects. The present paper is concerned with some problems regarding the behavior of synchronous BFDS in which the state update functions (or the local state transition functions) are chosen from a predetermined finite basis of boolean functions \\({\\mathcal {B}}\\). Specifically the following three behaviors are studied: Convergence. Does a system at hand converge on a given initial state configuration? Path Intersection. Will a system starting in given two state configurations produce a common configuration? Cycle Length. Since the state space is finite, every BFDS on a given initial state configuration either converges or enters a cycle having length greater than 1. If the latter is the case, what is the length of the loop? Or put more simply, for an integer \\(t\\), is the length of loop greater than \\(t\\)?

Parametric Wind Velocity Vector Estimation Method for Single Doppler LIDAR Model (Space, Aeronautical and Navigational Electronics)

Abstract: Doppler lidar (LIght Detection And Ranging) can provide accurate wind velocity vector estimates by processing the time delay and Doppler spectrum of received signals. This system is essential for realtime wind monitoring to assist aircraft taking off and landing. Considering the difficulty of calibration and cost, a single Doppler lidar model is more attractive and practical than a multiple lidar model. In general, it is impossible to estimate two or three dimensional wind vectors from a single lidar model without any prior information, because lidar directly observes only a 1-dimensional (radial direction) velocity component of wind. Although the conventional VAD (Velocity Azimuth Display) and VVP (Velocity Volume Processing) methods have been developed for single lidar model, both of them are inaccurate in the presence of local air turbulence. This paper proposes an accurate wind velocity estimation method based on a parametric approach using typical turbulence models such as tornado, micro-burst and gust front. The results from numerical simulation demonstrate that the proposed method remarkably enhances the accuracy for wind velocity estimation in the assumed modeled turbulence cases, compared with that obtained by the VAD or other conventional method. key words: Light Detection and Ranging (lidar), single lidar model, local air turbulence estimation