Papers by Arjun Mukherjee

5 papers
COIN – an Inexpensive and Strong Baseline for Predicting Out of Vocabulary Word Embeddings (2022.coling-1)

Copied to clipboard

Challenge: Word embedding models only include terms that occur a sufficient number of times in training corpora.
Approach: They propose a method for predicting word embeddings for out of vocabulary terms using word2vec.
Outcome: The proposed method surpasses several methods on benchmark tasks and is inexpensive to compute.
Experiments with Convolutional Neural Networks for Multi-Label Authorship Attribution (L18-1)

Copied to clipboard

Challenge: Existing methods for authorship attribution tasks are difficult, but they are effective.
Approach: They propose a CNN that averaging author probability distributions at sentence level for longer documents and treating smaller documents as sentences adapts to single-label datasets and various document sizes.
Outcome: The proposed method outperforms state-of-the-art models on a single-label AA benchmark dataset.
Attending Sentences to detect Satirical Fake News (C18-1)

Copied to clipboard

Challenge: Existing approaches to capture news satire do not explore sentence and document difference .
Approach: They propose a hierarchical deep neural network approach for satire detection . it is able to capture satirical news both at the sentence level and document level .
Outcome: The proposed approach can capture satire at sentence and document levels.
Predicting Personal Opinion on Future Events with Fingerprints (2020.coling-main)

Copied to clipboard

Challenge: Existing methods to predict users’ opinions on going events may not be able to acquire such content and thus cannot infer an unbiased opinion on emerging events.
Approach: They propose to model opinion on unseen articles based on one’s fingerprinting: the prior reading and commenting history.
Outcome: The proposed model can predict user’s opinion on unseen articles based on one’s fingerprinting: the prior reading and commenting history.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations