Papers by Fahime Same
A Linguistic Perspective on Reference: Choosing a Feature Set for Generating Referring Expressions in Context (2020.coling-main)
Copied to clipboard
| Challenge: | Various studies have raised the question of which factors play a role in the choice of referring expressions. |
| Approach: | They propose to use Feature Importance Ranking and Sequential Forward Search to evaluate the “importance” of features in machine learning algorithms for selecting the form of a referring expression in discourse context. |
| Outcome: | The proposed feature set includes 6 features from 4 classes, namely grammatical role, inherent features of the referent, antecedent form and recency. |
Constructing Distributions of Variation in Referring Expression Type from Corpora for Model Evaluation (2022.lrec-1)
Copied to clipboard
| Challenge: | referencing is a non-deterministic task, but the algorithms for RE generation are evaluated against corpora of written texts which only include one RE per reference. |
| Approach: | They propose a method for exploring variation in human RE choice on the basis of longitudinal corpora. |
| Outcome: | The proposed method shows agreement between the evaluations against human judgements and parallel evaluations. |
Experimental versus In-Corpus Variation in Referring Expression Choice (2024.lrec-main)
Copied to clipboard
| Challenge: | Against our expectations, the divergence is greatest between the corpus and the GPT model. |
| Approach: | They compare the results of three studies to examine how well the corpus can model variation . they find that experimental methodology introduces substantial noise . |
| Outcome: | The results show that the corpus can model variation captured from the corpuse and RE form choices made during experiments. |
Non-neural Models Matter: a Re-evaluation of Neural Referring Expression Generation Systems (2022.acl-long)
Copied to clipboard
| Challenge: | In recent years, neural models have outperformed rule-based and classic approaches in NLG. |
| Approach: | They evaluate two English datasets and evaluate their performance using automatic and human evaluations. |
| Outcome: | The proposed model outperforms rule-based and classic approaches on two English datasets and is compared with human-based models. |
Intrinsic Task-based Evaluation for Referring Expression Generation (2024.acl-long)
Copied to clipboard
| Challenge: | Referring Expression Generation (REG) models generate referring expressions that refer to referents at different points in a discourse. |
| Approach: | They propose to use a purely ratings-based human evaluation to evaluate REG models by completing two meta-level tasks. |
| Outcome: | The proposed evaluation makes the models more reliable and discriminable, and improves the quality of the REs. |