A Framework for Enhancing Word Embeddings with Task-Specific Information

2018 ◽  
Author(s):  
Xuan Zhang ◽  
Chunhong Zhang ◽  
Chenchen Guo ◽  
Yang Ji
Keyword(s):  
Specific Information ◽  
Word Embeddings

Towards syntax-aware token embeddings

2020 ◽  
pp. 1-30
Author(s):  
Diana Nicoleta Popa ◽  
Julien Perez ◽  
James Henderson ◽  
Eric Gaussier
Keyword(s):  
Deep Learning ◽  
Learning Algorithm ◽  
Fine Tuning ◽  
Linguistic Knowledge ◽  
Specific Information ◽  
Large Set ◽  
Word Embeddings ◽  
Learning Models ◽  
Lexical Resources ◽  
Arbitrary Graph

Abstract Distributional semantic word representations are at the basis of most modern NLP systems. Their usefulness has been proven across various tasks, particularly as inputs to deep learning models. Beyond that, much work investigated fine-tuning the generic word embeddings to leverage linguistic knowledge from large lexical resources. Some work investigated context-dependent word token embeddings motivated by word sense disambiguation, using sequential context and large lexical resources. More recently, acknowledging the need for an in-context representation of words, some work leveraged information derived from language modelling and large amounts of data to induce contextualised representations. In this paper, we investigate Syntax-Aware word Token Embeddings (SATokE) as a way to explicitly encode specific information derived from the linguistic analysis of a sentence in vectors which are input to a deep learning model. We propose an efficient unsupervised learning algorithm based on tensor factorisation for computing these token embeddings given an arbitrary graph of linguistic structure. Applying this method to syntactic dependency structures, we investigate the usefulness of such token representations as part of deep learning models of text understanding. We encode a sentence either by learning embeddings for its tokens and the relations between them from scratch or by leveraging pre-trained relation embeddings to infer token representations. Given sufficient data, the former is slightly more accurate than the latter, yet both provide more informative token embeddings than standard word representations, even when the word representations have been learned on the same type of context from larger corpora (namely pre-trained dependency-based word embeddings). We use a large set of supervised tasks and two major deep learning families of models for sentence understanding to evaluate our proposal. We empirically demonstrate the superiority of the token representations compared to popular distributional representations of words for various sentence and sentence pair classification tasks.

A Color Coded STEM/SEM Image Storage System

1981 ◽  
Vol 39 ◽  
pp. 272-273
Author(s):  
Y. Kokubo ◽  
W. H. Hardy ◽  
J. Dance ◽  
K. Jones
Keyword(s):  
Image Processing ◽  
Electron Beam ◽  
Storage System ◽  
Particle Analysis ◽  
Specific Information ◽  
X Rays ◽  
Sem Image ◽  
Backscattered Electrons ◽  
Wide Range ◽  
Scanning Electron

A color coded digital image processing is accomplished by using JEM100CX TEM SCAN and ORTEC’s LSI-11 computer based multi-channel analyzer (EEDS-II-System III) for image analysis and display. Color coding of the recorded image enables enhanced visualization of the image using mathematical techniques such as compression, gray scale expansion, gamma-processing, filtering, etc., without subjecting the sample to further electron beam irradiation once images have been stored in the memory.The powerful combination between a scanning electron microscope and computer is starting to be widely used 1) - 4) for the purpose of image processing and particle analysis. Especially, in scanning electron microscopy it is possible to get all information resulting from the interactions between the electron beam and specimen materials, by using different detectors for signals such as secondary electron, backscattered electrons, elastic scattered electrons, inelastic scattered electrons, un-scattered electrons, X-rays, etc., each of which contains specific information arising from their physical origin, study of a wide range of effects becomes possible.

Electron Microscopy in the Study of Natural Phytoplankton Assemblages

1985 ◽  
Vol 43 ◽  
pp. 620-623
Author(s):  
Linda Sicko-Goad
Keyword(s):  
Electron Microscopy ◽  
Aquatic Environment ◽  
Size Range ◽  
Physical Parameters ◽  
Specific Information ◽  
Phytoplankton Assemblages ◽  
Phytoplankton Productivity ◽  
Ecosystem Effects ◽  
Time Of Year ◽  
Species Specific

Although the use of electron microscopy and its varied methodologies is not usually associated with ecological studies, the types of species specific information that can be generated by these techniques are often quite useful in predicting long-term ecosystem effects. The utility of these techniques is especially apparent when one considers both the size range of particles found in the aquatic environment and the complexity of the phytoplankton assemblages.The size range and character of organisms found in the aquatic environment are dependent upon a variety of physical parameters that include sampling depth, location, and time of year. In the winter months, all the Laurentian Great Lakes are uniformly mixed and homothermous in the range of 1.1 to 1.7°C. During this time phytoplankton productivity is quite low.

Interindividual Differences in Cognitive Functioning Are Associated with Autobiographical Memory Retrieval Specificity in Older Adults

GeroPsych ◽  
2020 ◽  
Vol 33 (1) ◽  
pp. 15-29 ◽  
Author(s):  
Sarah Peters ◽  
Signy Sheldon
Keyword(s):  
Older Adults ◽  
Cognitive Functioning ◽  
Autobiographical Memory ◽  
Memory Retrieval ◽  
Cognitive Status ◽  
Specific Information ◽  
Generation Task ◽  
Interindividual Differences ◽  
Retrieval Cues ◽  
Contextual Support

Abstract. We examined whether interindividual differences in cognitive functioning among older adults are related to episodic memory engagement during autobiographical memory retrieval. Older adults ( n = 49, 24 males; mean age = 69.93; mean education = 15.45) with different levels of cognitive functioning, estimated using the Montreal Cognitive Assessment (MoCA), retrieved multiple memories (generation task) and the details of a single memory (elaboration task) to cues representing thematic or event-specific autobiographical knowledge. We found that the MoCA score positively predicted the proportion of specific memories for generation and episodic details for elaboration, but only to cues that represented event-specific information. The results demonstrate that individuals with healthy, but not unhealthy, cognitive status can leverage contextual support from retrieval cues to improve autobiographical specificity.

The Influence of Testing on Relational and Item-Specific Information

2013 ◽  
Author(s):  
Jocelyn Breton ◽  
Samantha Wasserman ◽  
Hilary Woodworth ◽  
Julianne Wieboldt ◽  
Hannah Newman ◽  
...  
Keyword(s):  
Specific Information

Tailoring Communications for Primary Care Settings

1998 ◽  
Vol 37 (02) ◽  
pp. 171-178 ◽  
Author(s):  
B. Glassman ◽  
B. K. Rimer
Keyword(s):  
Primary Care ◽  
Health Professionals ◽  
Contact Time ◽  
Behavioral Changes ◽  
Patient Specific ◽  
Specific Information ◽  
Medical Settings ◽  
Primary Care Settings ◽  
Professional Contact ◽  
Tailored Communications

AbstractIn more and more medical settings, physicians have less and less time to be effective communicators. To be effective, they need accurate, current information about their patients. Tailored health communications can facilitate positive patient-provider communications and foster behavioral changes conducive to health. Tailored communications (TCs) are produced for an individual based on information about that person. The focus of this report is on tailored print communications (TPCs). TPCs also enhance the process of evaluation, because they require a database and the collection of patient-specific information. We present a Tailoring Model for Primary Care that describes the steps involved in creating TPCs. We also provide examples from three ongoing studies in which TPCs are being used in order to illustrate the kinds of variables used for tailoring the products that are developed and how evaluation is conducted. TPCs offer opportunities to expand the reach of health professionals and to give personalized, individualized massages in an era of shrinking professional contact time.

Worker Protection Standard: Information at a Central Location

EDIS ◽  
2016 ◽  
Vol 2016 (7) ◽  
Author(s):  
Frederick M. Fishel ◽  
Tatiana Sanchez
Keyword(s):  
Direct Contact ◽  
Agricultural Workers ◽  
Specific Information ◽  
Federal Regulation ◽  
Central Location ◽  
Agricultural Pesticides ◽  
Full Implementation ◽  
Standard Information ◽  
Agricultural Plants ◽  
General Guidance

The Worker Protection Standard (WPS) is a Federal regulation designed to protect agricultural workers (people involved in the production of agricultural plants) and pesticide handlers (people mixing, loading, or applying pesticides or doing other tasks involving direct contact with pesticides). It has been in full implementation since 1995. This document provides general guidance to help you comply with the requirements of the Federal Worker Protection Standard (WPS) for agricultural pesticides, 40 CFR part 170, as amended in 2015. This document addresses specific information that agricultural establishments must display at a central location for their employees.

Sign in / Sign up

Export Citation Format

Share Document