Skip to content

Commit

Permalink
Add new task: Capturing discriminative attributes (SemEval 2018 Task …
Browse files Browse the repository at this point in the history 
…10) (sebastianruder#343)

* Add SemEval 2018 Task 10 section

* Sort section headers in alphabetical order
  • Loading branch information
@Yermouth @sebastianruder
Yermouth authored and sebastianruder committed Sep 15, 2019
1 parent a920eb1 commit 8606073
Showing 1 changed file with 59 additions and 34 deletions.
93 changes: 59 additions & 34 deletions english/relationship_extraction.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -4,41 +4,43 @@ Relationship extraction is the task of extracting semantic relationships from a
occur between two or more entities of a certain type (e.g. Person, Organisation, Location) and fall into a number of
semantic categories (e.g. married to, employed by, lives in).

### New York Times Corpus
### Capturing discriminative attributes (SemEval 2018 Task 10)

The standard corpus for distantly supervised relationship extraction is the New York Times (NYT) corpus, published in
[Riedel et al, 2010](http://www.riedelcastro.org//publications/papers/riedel10modeling.pdf).
**Capturing discriminative attributes (SemEval 2018 Task 10)** is a binary classification task where participants were asked to identify whether an attribute could help discriminate between two concepts. Unlike other word similarity prediction tasks, this task focuses on the semantic differences between words.

This contains text from the [New York Times Annotated Corpus](https://catalog.ldc.upenn.edu/ldc2008t19) with named
entities extracted from the text using the Stanford NER system and automatically linked to entities in the Freebase
knowledge base. Pairs of named entities are labelled with relationship types by aligning them against facts in the
Freebase knowledge base. (The process of using a separate database to provide label is known as 'distant supervision')
e.g. red(attribute) can be used to discriminate apple (concept1) from banana (concept2) -> label 1

Example:
> **Elevation Partners**, the $1.9 billion private equity group that was founded by **Roger McNamee**
More examples:
| Attribute | concept1 | concept2 | label |
| --------- | -------- | -------- | ----- |
| bookcase | fridge | wood | 1 |
| bucket | mug | round | 0 |
| angle | curve | sharp | 1 |
| pelican | turtle | water | 0 |
| wire | coil | metal | 0 |

`(founded_by, Elevation_Partners, Roger_McNamee)`
Task paper: [https://www.aclweb.org/anthology/S18-1117](https://www.aclweb.org/anthology/S18-1117)

Different papers have reported various metrics since the release of the dataset, making it difficult to compare systems
directly. The main metrics used are either precision at N results or plots of the precision-recall. The range of recall
has increased over the years as systems improve, with earlier systems having very low precision at 30% recall.
Task Codalab: [https://competitions.codalab.org/competitions/17326](https://competitions.codalab.org/competitions/17326)

| Model | Explainability | F1 Score | Paper / Source | Code |
| ----- | -------------- | -------- | -------------- | ---- |
| **SVM** with GloVe | **None** | **0.76** | [SUNNYNLP at SemEval-2018 Task 10: A Support-Vector-Machine-Based Method for Detecting Semantic Difference using Taxonomy and Word Embedding Features](https://aclweb.org/anthology/S18-1118) | [Author's](https://github.com/Yermouth/sunnynlp) |
| **SVM** with ConceptNet, Wikipedia articles and WordNet synonyms | None | 0.74 | [Luminoso at SemEval-2018 Task 10: Distinguishing Attributes Using Text Corpora and Relational Knowledge](https://aclweb.org/anthology/S18-1162) | [Author's](https://github.com/LuminosoInsight/semeval-discriminatt) |
| **MLP** combining information from various DSMs, PMI, and ConceptNet | None | 0.73 | [THU NGN at SemEval-2018 Task 10: Capturing Discriminative Attributes with MLP-CNN model](https://aclweb.org/anthology/S18-1157) | |
| **Gradient boosting** with co-occurrence count features and JoBimText features | None | 0.73 | [BomJi at SemEval-2018 Task 10: Combining Vector-, Pattern- and Graph-based Information to Identify Discriminative Attributes](https://aclweb.org/anthology/S18-1163) | |
| LexVec, word co-occurrence, and ConceptNet data combined using **maximum entropy classifier** | None | 0.72 | [UWB at SemEval-2018 Task 10: Capturing Discriminative Attributes from Word Distributions](https://aclweb.org/anthology/S18-1153) | [Author's](https://github.com/dpaperno/DiscriminAtt) |
| Composes explicit **vector spaces** from WordNet Definitions, ConceptNet and Visual Genome | **Fully Explainable** | **0.69** | [Identifying and Explaining Discriminative Attributes](https://arxiv.org/abs/1909.05363) | [Author's](https://github.com/ab-10/Hawk) |
| **Word2Vec** cosine similarities of WordNet glosses Transp. (No expl.) | Transp. (No expl.) | 0.69 | [Meaning space at SemEval-2018 Task 10: Combining explicitly encoded knowledge with information extracted from word embeddings](https://aclweb.org/anthology/S18-1154) | [Author's](https://github.com/cltl/meaning_space) |
| Use of Wikipedia and ConceptNet Transp. (No expl.) | Transp. (No expl.) | 0.69 | [ELiRF-UPV at SemEval-2018 Task 10: Capturing Discriminative Attributes with Knowledge Graphs and Wikipedia](https://aclweb.org/anthology/S18-1159) | |

| Model | P@10% | P@30% | Paper / Source | Code |
| ----------------------------------- | ----- | ----- | --------------- | -------------- |
| HRERE (Xu et al., 2019) | 84.9 | 72.8 | [Connecting Language and Knowledge with Heterogeneous Representations for Neural Relation Extraction](https://arxiv.org/abs/1903.10126) | [HRERE](https://github.com/billy-inn/HRERE) |
| PCNN+noise_convert+cond_opt (Wu et al., 2019) | 81.7 | 61.8 | [Improving Distantly Supervised Relation Extraction with Neural Noise Converter and Conditional Optimal Selector](https://arxiv.org/pdf/1811.05616.pdf) | |
| Intra- and Inter-Bag (Ye and Ling, 2019) | 78.9 | 62.4 | [Distant Supervision Relation Extraction with Intra-Bag and Inter-Bag Attentions](https://arxiv.org/pdf/1904.00143.pdf) | [Code](https://github.com/ZhixiuYe/Intra-Bag-and-Inter-Bag-Attentions) |
| RESIDE (Vashishth et al., 2018) | 73.6 | 59.5 | [RESIDE: Improving Distantly-Supervised Neural Relation Extraction using Side Information](http://malllabiisc.github.io/publications/papers/reside_emnlp18.pdf) | [RESIDE](https://github.com/malllabiisc/RESIDE) |
| PCNN+ATT (Lin et al., 2016) | 69.4 | 51.8 | [Neural Relation Extraction with Selective Attention over Instances](http://www.aclweb.org/anthology/P16-1200) | [OpenNRE](https://github.com/thunlp/OpenNRE/) |
| MIML-RE (Surdeneau et al., 2012) | 60.7+ | - | [Multi-instance Multi-label Learning for Relation Extraction](http://www.aclweb.org/anthology/D12-1042) | [Mimlre](https://nlp.stanford.edu/software/mimlre.shtml) |
| MultiR (Hoffman et al., 2011) | 60.9+ | - | [Knowledge-Based Weak Supervision for Information Extraction of Overlapping Relations](http://www.aclweb.org/anthology/P11-1055) | [MultiR](http://aiweb.cs.washington.edu/ai/raphaelh/mr/) |
| (Mintz et al., 2009) | 39.9+ | - | [Distant supervision for relation extraction without labeled data](http://www.aclweb.org/anthology/P09-1113) | |
### FewRel

The Few-Shot Relation Classification Dataset (FewRel) is a different setting from the previous datasets. This dataset consists of 70K sentences expressing 100 relations annotated by crowdworkers on Wikipedia corpus. The few-shot learning task follows the N-way K-shot meta learning setting. It is both the largest supervised relation classification dataset as well as the largest few-shot learning dataset till now.

(+) Obtained from results in the paper "Neural Relation Extraction with Selective Attention over Instances"
The public leaderboard is available on the [FewRel website](http://www.zhuhao.me/fewrel/).

### SemEval-2010 Task 8
### Multi-Way Classification of Semantic Relations Between Pairs of Nominals (SemEval 2010 Task 8)

[SemEval-2010](http://www.aclweb.org/anthology/S10-1006) introduced 'Task 8 - Multi-Way Classification of Semantic
Relations Between Pairs of Nominals'. The task is, given a sentence and two tagged nominals, to predict the relation
Expand Down Expand Up @@ -75,7 +77,6 @@ reported here are the highest achieved by the model using any external resources

<a name="footnote">*</a>: It uses external lexical resources, such as WordNet, part-of-speech tags, dependency tags, and named entity tags.


#### Dependency Models

| Model | F1 | Paper / Source | Code |
Expand All @@ -88,6 +89,38 @@ reported here are the highest achieved by the model using any external resources
| FCN (Yu et al., 2014) | 83.0 | [Factor-based compositional embedding models](https://www.cs.cmu.edu/~mgormley/papers/yu+gormley+dredze.nipsw.2014.pdf) |
| MVRNN (Socher et al., 2012) | 82.4 | [Semantic Compositionality through Recursive Matrix-Vector Spaces](http://aclweb.org/anthology/D12-1110) | [pratapbhanu's Reimplementation](https://github.com/pratapbhanu/MVRNN) |

### New York Times Corpus

The standard corpus for distantly supervised relationship extraction is the New York Times (NYT) corpus, published in
[Riedel et al, 2010](http://www.riedelcastro.org//publications/papers/riedel10modeling.pdf).

This contains text from the [New York Times Annotated Corpus](https://catalog.ldc.upenn.edu/ldc2008t19) with named
entities extracted from the text using the Stanford NER system and automatically linked to entities in the Freebase
knowledge base. Pairs of named entities are labelled with relationship types by aligning them against facts in the
Freebase knowledge base. (The process of using a separate database to provide label is known as 'distant supervision')

Example:
> **Elevation Partners**, the $1.9 billion private equity group that was founded by **Roger McNamee**
`(founded_by, Elevation_Partners, Roger_McNamee)`

Different papers have reported various metrics since the release of the dataset, making it difficult to compare systems
directly. The main metrics used are either precision at N results or plots of the precision-recall. The range of recall
has increased over the years as systems improve, with earlier systems having very low precision at 30% recall.


| Model | P@10% | P@30% | Paper / Source | Code |
| ----------------------------------- | ----- | ----- | --------------- | -------------- |
| HRERE (Xu et al., 2019) | 84.9 | 72.8 | [Connecting Language and Knowledge with Heterogeneous Representations for Neural Relation Extraction](https://arxiv.org/abs/1903.10126) | [HRERE](https://github.com/billy-inn/HRERE) |
| PCNN+noise_convert+cond_opt (Wu et al., 2019) | 81.7 | 61.8 | [Improving Distantly Supervised Relation Extraction with Neural Noise Converter and Conditional Optimal Selector](https://arxiv.org/pdf/1811.05616.pdf) | |
| Intra- and Inter-Bag (Ye and Ling, 2019) | 78.9 | 62.4 | [Distant Supervision Relation Extraction with Intra-Bag and Inter-Bag Attentions](https://arxiv.org/pdf/1904.00143.pdf) | [Code](https://github.com/ZhixiuYe/Intra-Bag-and-Inter-Bag-Attentions) |
| RESIDE (Vashishth et al., 2018) | 73.6 | 59.5 | [RESIDE: Improving Distantly-Supervised Neural Relation Extraction using Side Information](http://malllabiisc.github.io/publications/papers/reside_emnlp18.pdf) | [RESIDE](https://github.com/malllabiisc/RESIDE) |
| PCNN+ATT (Lin et al., 2016) | 69.4 | 51.8 | [Neural Relation Extraction with Selective Attention over Instances](http://www.aclweb.org/anthology/P16-1200) | [OpenNRE](https://github.com/thunlp/OpenNRE/) |
| MIML-RE (Surdeneau et al., 2012) | 60.7+ | - | [Multi-instance Multi-label Learning for Relation Extraction](http://www.aclweb.org/anthology/D12-1042) | [Mimlre](https://nlp.stanford.edu/software/mimlre.shtml) |
| MultiR (Hoffman et al., 2011) | 60.9+ | - | [Knowledge-Based Weak Supervision for Information Extraction of Overlapping Relations](http://www.aclweb.org/anthology/P11-1055) | [MultiR](http://aiweb.cs.washington.edu/ai/raphaelh/mr/) |
| (Mintz et al., 2009) | 39.9+ | - | [Distant supervision for relation extraction without labeled data](http://www.aclweb.org/anthology/P09-1113) | |

(+) Obtained from results in the paper "Neural Relation Extraction with Selective Attention over Instances"

### TACRED

Expand All @@ -106,12 +139,4 @@ _no_relation_ type).
| C-GCN + PA-LSTM (Zhang et al. 2018) | **68.2** | [Graph Convolution over Pruned Dependency Trees Improves Relation Extraction](http://aclweb.org/anthology/D18-1244) | [Offical](https://github.com/qipeng/gcn-over-pruned-trees) |
| PA-LSTM (Zhang et al, 2017) | 65.1 | [Position-aware Attention and Supervised Data Improve Slot Filling](http://aclweb.org/anthology/D17-1004) | [Official](https://github.com/yuhaozhang/tacred-relation) |



# FewRel

The Few-Shot Relation Classification Dataset (FewRel) is a different setting from the previous datasets. This dataset consists of 70K sentences expressing 100 relations annotated by crowdworkers on Wikipedia corpus. The few-shot learning task follows the N-way K-shot meta learning setting. It is both the largest supervised relation classification dataset as well as the largest few-shot learning dataset till now.

The public leaderboard is available on the [FewRel website](http://www.zhuhao.me/fewrel/).

[Go back to the README](../README.md)

0 comments on commit 8606073

Please sign in to comment.