Essentia: Mining Domain-specific Paraphrases with Word-Alignment Graphs (D19-53)
Copied to clipboard
| Challenge: | Existing methods for mining general-purpose paraphrases are often based on statistical methods, but domain-specific corpora are too small to fit statistical methods. |
| Approach: | They propose a method to mine paraphrases from a small set of sentences that roughly share the same topic or intent. |
| Outcome: | The proposed method obtains high quality paraphrases as evaluated by crowd workers. |
Similar Papers
A Large Resource of Patterns for Verbal Paraphrases (L18-1)
Copied to clipboard
| Challenge: | Xu et al., 2015: paraphrases play an important role in natural language understanding . he says it is difficult to propose a paraphrasing relation for natural language processing systems . |
| Approach: | They propose a resource of such paraphrases that can be used to identify hidden paraphrase pairs . they propose to use the resource to identify paraphrase relationships between two words . |
| Outcome: | The proposed resource contains tens of thousands of such pairs and is available for academic purposes. |
Using Paraphrases to Study Properties of Contextual Embeddings (2022.naacl-main)
Copied to clipboard
| Challenge: | Previously, paraphrases have been used to probe whether compositionality is accurately captured by BERT, but we believe they can be used to explore many other questions. |
| Approach: | They propose to use paraphrases as a unique source of data to analyze contextualized embeddings, with a particular focus on BERT. |
| Outcome: | The proposed analysis of paraphrases and paraphrase representations using the Paraphrase Database shows that BERT handles polysemous words, but different representations in many cases. |
SPADE: Evaluation Dataset for Monolingual Phrase Alignment (L18-1)
Copied to clipboard
| Challenge: | Existing studies on sentential paraphrase detection focus on finer grained paraphrases, i.e., phrasal paraphrase. |
| Approach: | They propose to use the SPADE to evaluate syntactic phrase alignment in paraphrasal sentences. |
| Outcome: | The proposed method is compared with humans and provides benchmarks to show its performance. |
Improving Paraphrase Detection with the Adversarial Paraphrasing Task (2021.acl-long)
Copied to clipboard
| Challenge: | a new adversarial method of paraphrase identification is being used to identify paraphrases based on word overlap and syntax . authors propose a dataset that generates semantically equivalent but lexically and syntactically disparate paraphrase pairs . |
| Approach: | They propose an adversarial method for paraphrase identification that uses word overlap and syntax to identify paraphrases. |
| Outcome: | The proposed method improves paraphrase detection accuracy and speed of generation of datasets. |
What’s Mine becomes Yours: Defining, Annotating and Detecting Context-Dependent Paraphrases in News Interview Dialogs (2024.emnlp-main)
Copied to clipboard
| Challenge: | a dataset of utterance pairs from NPR and CNN is used to classify paraphrases in dialog. |
| Approach: | They propose a dataset annotated for context-dependent paraphrases and develop a training for crowd-workers to classify paraphrase in dialog. |
| Outcome: | The proposed dataset contains 5,581 annotations on 600 utterance pairs. |
SaSAKE: Syntax and Semantics Aware Keyphrase Extraction from Research Papers (2020.coling-main)
Copied to clipboard
| Challenge: | Syntax and semantics Aware Keyphrase Extraction (SaSAKE) is a novel method for extracting keyphrases from research papers. |
| Approach: | They propose a method to automatically extract keyphrases from research papers using a transformer architecture and graph encoders to incorporate syntactic and semantic dependency graphs. |
| Outcome: | The proposed method achieves state-of-the-art performance on three benchmark datasets. |
Paraphrase Types for Generation and Detection (2023.emnlp-main)
Copied to clipboard
| Challenge: | Current approaches to paraphrase generation and detection ignore the intricate linguistic properties of language. |
| Approach: | They propose two tasks to consider specific linguistic perturbations at particular text positions. |
| Outcome: | The proposed tasks address the shortcoming of ignoring the linguistic properties of language. |
Text Mining for History: first steps on building a large dataset (L18-1)
Copied to clipboard
| Challenge: | a new corpus on the history domain is being created to mine text in the domain . primary motivation for the project is the need to query the material in a non-linear way . |
| Approach: | They propose to use a Brazilian historical-biographical dictionary as a resource for text mining. |
| Outcome: | The proposed corpus is a reference work on the Brazilian history domain . it contains almost 12 millions tokens in about three hundred thousand sentences . the authors argue that the proposed corpu is linguistically motivated . |
A Survey on Recent Advances in Keyphrase Extraction from Pre-trained Language Models (2023.findings-eacl)
Copied to clipboard
| Challenge: | Keyphrase extraction is a key component in Natural Language Processing (NLP) systems for selecting a set of phrases from the document that could summarize the important information discussed in the source document. |
| Approach: | They propose to use supervised and unsupervised keyphrase extraction techniques to investigate the state-of-the-art models for keyphrase extracting. |
| Outcome: | The proposed keyphrase extraction system can significantly accelerate the speed of retrieval and help people get first-hand information from a long document quickly and accurately. |
Exploiting Position and Contextual Word Embeddings for Keyphrase Extraction from Scientific Papers (2021.eacl-main)
Copied to clipboard
| Challenge: | Existing methods for keyphrase extraction are either supervised or unsupervised. |
| Approach: | They propose an unsupervised algorithm that exploits contextual word embeddings and positional information to create a biased PageRank. |
| Outcome: | The proposed algorithm outperforms previous approaches and strong baselines on five benchmark datasets. |