Xlm-v: Overcoming the vocabulary bottleneck in multilingual masked language models

D Liang, H Gonen, Y Mao, R Hou, N Goyal… - ar** a machine learning
classifier that can track high quality written feedback during peer review sessions. The …

A study of public sentiment and influence of politics in COVID-19 related tweets

A Mukherjee, S Bhattacharyya, K Ray, B Gupta… - … Intelligence in Pattern …, 2022 - Springer
An increased usage of social media was observed as the Covid-19 pandemic progressed.
We extract tweets relevant to the pandemic from publicly available Twitter data and analyse …

Sentence analogies: Exploring linguistic relationships and regularities in sentence embeddings

X Zhu, G de Melo - arxiv preprint arxiv:2003.04036, 2020 - arxiv.org
While important properties of word vector representations have been studied extensively, far
less is known about the properties of sentence vector representations. Word vectors are …

[KSIĄŻKA][B] Validity Challenges in Machine Learning Benchmarks

JP Miller - 2022 - search.proquest.com
Over the last decade, machine learning practitioners in fields like computer vision and
natural language processing have devoted vast resources to building models that …

Five Shades of BERT: Assessing Performance in Question Answering through Comparative Analysis

K Raikwar, P Rajusth, S Prabhu… - 2024 4th Asian …, 2024 - ieeexplore.ieee.org
The Bidirectional Encoder Representations from Transformers (BERT) model has
transformed natural language processing, achieving great results across a wide range of …