Iker commited on
Commit
98355e3
1 Parent(s): 89e5f10

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +52 -0
README.md CHANGED
@@ -51,10 +51,62 @@ size_categories:
51
 
52
  ---
53
 
 
 
 
 
 
 
 
54
  ```python
55
  {
56
  "O": 0,
57
  "B-TARGET": 1,
58
  "I-TARGET": 2
59
  }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
60
  ```
 
51
 
52
  ---
53
 
54
+ This repository contains the English '[SemEval-2014 Task 4: Aspect Based Sentiment Analysis](https://aclanthology.org/S14-2004/)'. translated with DeepL into Spanish, French, Russian, and Turkish. The **labels have been manually projected**. For more details, read this paper: [Model and Data Transfer for Cross-Lingual Sequence Labelling in Zero-Resource Settings](https://arxiv.org/abs/2210.12623).
55
+
56
+ **Intended Usage**: Since the datasets are parallel across languages, they are ideal for evaluating annotation projection algorithms, such as [T-Projection](https://arxiv.org/abs/2212.10548).
57
+
58
+
59
+ # Label Dictionary
60
+
61
  ```python
62
  {
63
  "O": 0,
64
  "B-TARGET": 1,
65
  "I-TARGET": 2
66
  }
67
+ ```
68
+
69
+ # Cication
70
+
71
+ If you use this data, please cite the following papers:
72
+
73
+ ```bibtex
74
+ @inproceedings{garcia-ferrero-etal-2022-model,
75
+ title = "Model and Data Transfer for Cross-Lingual Sequence Labelling in Zero-Resource Settings",
76
+ author = "Garc{\'\i}a-Ferrero, Iker and
77
+ Agerri, Rodrigo and
78
+ Rigau, German",
79
+ editor = "Goldberg, Yoav and
80
+ Kozareva, Zornitsa and
81
+ Zhang, Yue",
82
+ booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2022",
83
+ month = dec,
84
+ year = "2022",
85
+ address = "Abu Dhabi, United Arab Emirates",
86
+ publisher = "Association for Computational Linguistics",
87
+ url = "https://aclanthology.org/2022.findings-emnlp.478",
88
+ doi = "10.18653/v1/2022.findings-emnlp.478",
89
+ pages = "6403--6416",
90
+ abstract = "Zero-resource cross-lingual transfer approaches aim to apply supervised modelsfrom a source language to unlabelled target languages. In this paper we performan in-depth study of the two main techniques employed so far for cross-lingualzero-resource sequence labelling, based either on data or model transfer. Although previous research has proposed translation and annotation projection(data-based cross-lingual transfer) as an effective technique for cross-lingualsequence labelling, in this paper we experimentally demonstrate that highcapacity multilingual language models applied in a zero-shot (model-basedcross-lingual transfer) setting consistently outperform data-basedcross-lingual transfer approaches. A detailed analysis of our results suggeststhat this might be due to important differences in language use. Morespecifically, machine translation often generates a textual signal which isdifferent to what the models are exposed to when using gold standard data,which affects both the fine-tuning and evaluation processes. Our results alsoindicate that data-based cross-lingual transfer approaches remain a competitiveoption when high-capacity multilingual language models are not available.",
91
+ }
92
+
93
+ @inproceedings{pontiki-etal-2014-semeval,
94
+ title = "{S}em{E}val-2014 Task 4: Aspect Based Sentiment Analysis",
95
+ author = "Pontiki, Maria and
96
+ Galanis, Dimitris and
97
+ Pavlopoulos, John and
98
+ Papageorgiou, Harris and
99
+ Androutsopoulos, Ion and
100
+ Manandhar, Suresh",
101
+ editor = "Nakov, Preslav and
102
+ Zesch, Torsten",
103
+ booktitle = "Proceedings of the 8th International Workshop on Semantic Evaluation ({S}em{E}val 2014)",
104
+ month = aug,
105
+ year = "2014",
106
+ address = "Dublin, Ireland",
107
+ publisher = "Association for Computational Linguistics",
108
+ url = "https://aclanthology.org/S14-2004",
109
+ doi = "10.3115/v1/S14-2004",
110
+ pages = "27--35",
111
+ }
112
  ```