File size: 9,278 Bytes
38beee7
 
 
 
 
 
d387681
447e869
38beee7
 
 
 
 
e673247
 
d2b926f
e673247
d2b926f
2bb8b4d
d2b926f
2bb8b4d
447e869
 
 
 
 
38beee7
 
 
d387681
38beee7
 
 
d387681
 
 
 
 
 
 
38beee7
 
d387681
38beee7
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
---
license: cc-by-4.0
language: 
- pl
- en
datasets:
- posmac
pipeline_tag: text2text-generation
tags:
- keywords-generation
- text-classifiation
- other
widget:
- text: "Keywords: Our vlT5 model is a keyword generation model based on encoder-decoder architecture using Transformer blocks presented by google (https://huggingface.co/t5-base). The vlT5 was trained on scientific articles corpus to predict a given set of keyphrases based on the concatenation of the article’s abstract and title. It generates precise, yet not always complete keyphrases that describe the content of the article based only on the abstract."
  example_title: "English 1"
- text: "Keywords: Decays the learning rate of each parameter group by gamma every step_size epochs. Notice that such decay can happen simultaneously with other changes to the learning rate from outside this scheduler. When last_epoch=-1, sets initial lr as lr."
  example_title: "English 2"
- text: "Keywords: Przełomem w dziedzinie sztucznej inteligencji i maszynowego uczenia się było powstanie systemu eksperckiego Dendral na Uniwersytecie Stanforda w 1965. System ten powstał w celu zautomatyzowania analizy i identyfikacji molekuł związków organicznych, które dotychczas nie były znane chemikom. Wyniki badań otrzymane dzięki systemowi Dendral były pierwszym w historii odkryciem dokonanym przez komputer, które zostały opublikowane w prasie specjalistycznej."
  example_title: "Polish"
- text: "Keywords: El Padrão real (traducible al español como Patrón real) era una obra cartográfica de origen portugués producida secretamente y mantenida por la organización de la corte real en el siglo XVI. La obra estaba disponible para la élite científica de la época, siendo expuesta en la Casa da Índia (Casa de la India). En el Padrão real se añadieron constantemente los nuevos descubrimientos de los portugueses. El primer Padrão real fue producido en la época de Enrique el Navegante, antes de la existencia de la Casa de la India. "
  example_title: "Spanish"
metrics:
- f1
- precision
- recall

---
# Keyword Extraction from Short Texts with T5

Our vlT5 model is a keyword generation model based on encoder-decoder architecture using Transformer blocks presented by Google ([https://huggingface.co/t5-base](https://huggingface.co/t5-base)). The vlT5 was trained on scientific articles corpus to predict a given set of keyphrases based on the concatenation of the article’s abstract and title. It generates precise, yet not always complete keyphrases that describe the content of the article based only on the abstract.

The biggest advantage is the transferability of the vlT5 model, as it works well on all domains and types of text. The downside is that the text length and the number of keywords are similar to the training data: the text piece of an abstract length generates approximately 3 to 5 keywords. It works both extractive and abstractively. Longer pieces of text must be split into smaller chunks, and then propagated to the model.

### Overview
- **Language model:** [t5-base](https://huggingface.co/t5-base)   
- **Language:** pl, en (but works relatively well with others)
- **Training data:** POSMAC
- **Online Demo:** [https://nlp-demo-1.voicelab.ai/](https://nlp-demo-1.voicelab.ai/)
- **Paper:** [TBA](TBA)

# Corpus

The model was trained on a POSMAC corpus. Polish Open Science Metadata Corpus (POSMAC) is a collection of  216,214 abstracts of scientific publications compiled in the CURLICAT project.


| Domains                                                  | Documents | With keywords |
| -------------------------------------------------------- | --------: | :-----------: |
| Engineering and technical sciences                       |    58 974 |    57 165    |
| Social sciences                                          |    58 166 |    41 799    |
| Agricultural sciences                                    |    29 811 |    15 492    |
| Humanities                                               |    22 755 |    11 497    |
| Exact and natural sciences                               |    13 579 |     9 185     |
| Humanities, Social sciences                              |    12 809 |     7 063     |
| Medical and health sciences                              |     6 030 |     3 913     |
| Medical and health sciences, Social sciences             |       828 |      571      |
| Humanities, Medical and health sciences, Social sciences |       601 |      455      |
| Engineering and technical sciences, Humanities           |       312 |      312      |

# Tokenizer

As in the original HerBERT implementation, the training dataset was tokenized into subwords using a character level byte-pair encoding (CharBPETokenizer) with a vocabulary size of 50k tokens. The tokenizer itself was trained with a tokenizers library.

We kindly encourage you to use the Fast version of the tokenizer, namely HerbertTokenizerFast.

# Usage

```python
from transformers import T5Tokenizer, T5ForConditionalGeneration

vlt5 = T5ForConditionalGeneration.from_pretrained("Voicelab/t5-base-keywords")
tokenizer = T5Tokenizer.from_pretrained("Voicelab/t5-base-keywords")

task_prefix = "Keywords: "
inputs = ["Christina Katrakis, who spoke to the BBC from Vorokhta in western Ukraine, relays the account of one family, who say Russian soldiers shot at their vehicles while they were leaving their village near Chernobyl in northern Ukraine. She says the cars had white flags and signs saying they were carrying children.",
     "Decays the learning rate of each parameter group by gamma every step_size epochs. Notice that such decay can happen simultaneously with other changes to the learning rate from outside this scheduler. When last_epoch=-1, sets initial lr as lr.",
    "Hello, I'd like to order a pizza with salami topping."]

for sample in inputs:
       	input_sequences = [task_prefix + sample]
        input_ids = tokenizer(input_sequences, return_tensors='pt', truncation=True).input_ids
        output = model.generate(input_ids, no_repeat_ngram_size=3, num_beams=4)
        predicted = tokenizer.decode(output[0], skip_special_tokens=True)
        print(sample, "\n --->", predicted)


  
```

# Results



| Method      | Rank |   Micro   |            |            | Macro |       |       |
| ----------- | ---: | :--------: | ---------: | ---------: | :---: | ----: | ----: |
|             |      |     P     |          R |         F1 |   P   |     R |    F1 |
| extremeText |    1 |   0.175   |      0.038 |      0.063 | 0.007 | 0.004 | 0.005 |
|             |    3 |   0.117   |      0.077 |      0.093 | 0.011 | 0.011 | 0.011 |
|             |    5 |   0.090   |      0.099 |      0.094 | 0.013 | 0.016 | 0.015 |
|             |   10 |   0.060   |      0.131 |      0.082 | 0.015 | 0.025 | 0.019 |
| plT5kw      |    1 | **0.345** |      0.076 |      0.124 | 0.054 | 0.047 | 0.050 |
|             |    3 |   0.328   |      0.212 |      0.257 | 0.133 | 0.127 | 0.129 |
|             |    5 |   0.318   | **0.237** | **0.271** | 0.143 | 0.140 | 0.141 |
| KeyBERT     |    1 |   0.030   |      0.007 |      0.011 | 0.004 | 0.003 | 0.003 |
|             |    3 |   0.015   |      0.010 |      0.012 | 0.006 | 0.004 | 0.005 |
|             |    5 |   0.011   |      0.012 |      0.011 | 0.006 | 0.005 | 0.005 |
| TermoPL     |    1 |   0.118   |      0.026 |      0.043 | 0.004 | 0.003 | 0.003 |
|             |    3 |   0.070   |      0.046 |      0.056 | 0.006 | 0.005 | 0.006 |
|             |    5 |   0.051   |      0.056 |      0.053 | 0.007 | 0.007 | 0.007 |
|             |  all |   0.025   |      0.339 |      0.047 | 0.017 | 0.030 | 0.022 |
| extremeText |    1 |   0.210   |      0.077 |      0.112 | 0.037 | 0.017 | 0.023 |
|             |    3 |   0.139   |      0.152 |      0.145 | 0.045 | 0.042 | 0.043 |
|             |    5 |   0.107   |      0.196 |      0.139 | 0.049 | 0.063 | 0.055 |
|             |   10 |   0.072   |      0.262 |      0.112 | 0.041 | 0.098 | 0.058 |
| plT5kw      |    1 | **0.377** |      0.138 |      0.202 | 0.119 | 0.071 | 0.089 |
|             |    3 |   0.361   |      0.301 |      0.328 | 0.185 | 0.147 | 0.164 |
|             |    5 |   0.357   | **0.316** | **0.335** | 0.188 | 0.153 | 0.169 |
| KeyBERT     |    1 |   0.018   |      0.007 |      0.010 | 0.003 | 0.001 | 0.001 |
|             |    3 |   0.009   |      0.010 |      0.009 | 0.004 | 0.001 | 0.002 |
|             |    5 |   0.007   |      0.012 |      0.009 | 0.004 | 0.001 | 0.002 |
| TermoPL     |    1 |   0.076   |      0.028 |      0.041 | 0.002 | 0.001 | 0.001 |
|             |    3 |   0.046   |      0.051 |      0.048 | 0.003 | 0.001 | 0.002 |
|             |    5 |   0.033   |      0.061 |      0.043 | 0.003 | 0.001 | 0.002 |
|             |  all |   0.021   |      0.457 |      0.040 | 0.004 | 0.008 | 0.005 |

# License

CC BY 4.0

# Citation

If you use this model, please cite the following paper:

Piotr Pęzik, Agnieszka Mikołajczyk-Bareła, Adam Wawrzyński, Bartłomiej Nitoń, Maciej Ogrodniczuk, Keyword Extraction from Short Texts with a Text-To-Text Transfer Transformer, ACIIDS 2022


# Authors

The model was trained by NLP Research Team at Voicelab.ai.

You can contact us [here](https://voicelab.ai/contact/).