Papers with Deep neural network models
Affection Driven Neural Networks for Sentiment Analysis (2020.lrec-1)
Copied to clipboard
| Challenge: | Existing deep neural network models lack mechanisms to highlight important sentiment terms. |
| Approach: | They propose a method to incorporate affective knowledge into deep neural network models by mapping affective influence vectors to an affective impact value and integrating them into long-term memory models to highlight affective terms. |
| Outcome: | The proposed approach improves on three large datasets by 1.0% to 1.5% on the benchmark datasets. |
EcoSpeak: Cost-Efficient Bias Mitigation for Partially Cross-Lingual Speaker Verification (2024.findings-naacl)
Copied to clipboard
| Challenge: | Linguistic bias is a critical problem that harms the diversity, equity, and inclusiveness of Natural Language Processing tools. |
| Approach: | They propose a low-cost solution that incorporates contrastive linguistic attention to emphasize relevant speaker verification embedding parts. |
| Outcome: | The proposed model is able to mitigate linguistic bias in five partially cross-lingual scenarios. |
Investigating Dynamic Routing in Tree-Structured LSTM for Sentiment Analysis (D19-1)
Copied to clipboard
| Challenge: | Existing deep neural network models such as LSTM and tree-LSTM have a bias problem where the words in the tail of a sentence are more heavily emphasized than those in the header. |
| Approach: | They propose a capsule tree-LSTM model that uses dynamic routing to build sentence representations by assigning different weights to nodes according to their contributions to prediction. |
| Outcome: | The proposed model improves on the Stanford Sentiment Treebank and EmoBank datasets. |
Named Entity Recognition Only from Word Embeddings (2020.emnlp-main)
Copied to clipboard
| Challenge: | Existing named entity recognition systems require large amounts of human annotated training data. |
| Approach: | They propose a fully unsupervised named entity recognition model which takes clues from pre-trained word embeddings. |
| Outcome: | The proposed model can be trained on two CoNLL benchmark datasets without annotating lexicon or corpus. |