nlp   30686

« earlier    

And if this works for natural languages ... why not trying programming languages next?
NLP  MLonCode  from twitter_favs
17 hours ago by ngpestelos
[1803.11175] Universal Sentence Encoder
We present models for encoding sentences into embedding vectors that specifically target transfer learning to other NLP tasks. The models are efficient and result in accurate performance on diverse transfer tasks. Two variants of the encoding models allow for trade-offs between accuracy and compute resources. For both variants, we investigate and report the relationship between model complexity, resource consumption, the availability of transfer task training data, and task performance. Comparisons are made with baselines that use word level transfer learning via pretrained word embeddings as well as baselines do not use any transfer learning. We find that transfer learning using sentence embeddings tends to outperform word level transfer. With transfer learning via sentence embeddings, we observe surprisingly good performance with minimal amounts of supervised training data for a transfer task. We obtain encouraging results on Word Embedding Association Tests (WEAT) targeted at detecting model bias. Our pre-trained sentence encoding models are made freely available for download and on TF Hub.
nlp  machine_learning 
yesterday by amy
[1804.07754] Learning Semantic Textual Similarity from Conversations
We present a novel approach to learn representations for sentence-level semantic similarity using conversational data. Our method trains an unsupervised model to predict conversational input-response pairs. The resulting sentence embeddings perform well on the semantic textual similarity (STS) benchmark and SemEval 2017's Community Question Answering (CQA) question similarity subtask. Performance is further improved by introducing multitask training combining the conversational input-response prediction task and a natural language inference task. Extensive experiments show the proposed model achieves the best performance among all neural models on the STS benchmark and is competitive with the state-of-the-art feature engineered and mixed systems in both tasks.
nlp  machine_learning 
yesterday by amy
Comparing Brazilian and US university theses using natural language processing
Although the subject matter doesn't interest me the process could be applied to other areas and provides a "beginners guide" to a certain type of text analysis.
ai  nlp 
yesterday by shearichard
ciprian-chelba/1-billion-word-language-modeling-benchmark: Formerly known as
GitHub is where people build software. More than 27 million people use GitHub to discover, fork, and contribute to over 80 million projects.
lm1b  data  dataset  download  nlp  deep-learning 
yesterday by nharbour

« earlier    

related tags

ai  algorithms  analysis  ann  annotation  arabic  archive  artificialintelligence  arxiv  automation  awesome  aws  azure  best-of-2018  bmi  books  bot  business  c  chainer  chatbots  classification  collections  comprehension  concurrency  conference  cool  corpus  correct  correction  data-mining  data-visualization  data  data_science  datascience  dataset  datasets  deep-learning  deeplearn  deeplearning  dh  dialog  dl  download  embeddings  evaluations  fastai  fasttext  free  generators  geo  gigaword  github  gluon  gmail  google  history  hn  humor  ideas  intelligence  investing  ios  javascript  jupyter-notebook  jupyter  keras  labeling  labelling  labels  language  law  lda  libraries  linguistics  linked-data  lm1b  lstm  machine-learing  machine-learning  machine_learning  machinelearning  ml  mloncode  movie  mxnet  natural-language  ner  neuralnetworks  nlg  nodejs  notebook  onlinetools  pandas  paper  people  pointer-sentinel  pointer  politeness  programming  python  pytorch  reading  research  researchers  rnn  rpa  ruder-newsletter  search  sentiment  sentinel  simone-teufel  space  spacy  spell-check  spelling  split  splitting  swiss  text  textdata  texts  tool  torchtext  transfer-learning  transfer_learning  transferlearning  tutorial  typos  uk  unicode  visualisation  vsm  wikipedia  word2vec  workshop  writing 

Copy this bookmark: