Papers by Hassan Hajipoor
Serial Recall Effects in Neural Language Modeling (N19-1)
Copied to clipboard
| Challenge: | Recent studies have shed light on the information encoded by LSTM networks. |
| Approach: | They propose to use serial recall experiments to model human memory of words in the order they occur in the language. |
| Outcome: | The proposed model can learn function words much better than content words and can capture syntactic structures such as subject-verb agreement. |
Embedding Time Differences in Context-sensitive Neural Networks for Learning Time to Event (2021.acl-short)
Copied to clipboard
| Challenge: | Current approaches focus on news articles and expect at least one temporal expressions in each input data to predict TTE. |
| Approach: | They propose a context-sensitive neural model for time to event prediction task . they enrich the model with time difference embeddings to improve accuracy . |
| Outcome: | The proposed model is 1.4 and 3.3 hours more accurate than the current state-of-the-art model on English and Dutch tweets respectively. |