Class: Informers::NER

Inherits:
Object
  • Object
show all
Defined in:
lib/informers/ner.rb

Instance Method Summary collapse

Constructor Details

#initialize(model_path) ⇒ NER

Returns a new instance of NER.



18
19
20
21
22
 
# File 'lib/informers/ner.rb', line 18

def initialize(model_path)
  tokenizer_path = File.expand_path("../../vendor/bert_base_cased_tok.bin", __dir__)
  @tokenizer = BlingFire.load_model(tokenizer_path)
  @model = OnnxRuntime::Model.new(model_path)
end

Instance Method Details

#predict(texts) ⇒ Object 



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
 
# File 'lib/informers/ner.rb', line 24

def predict(texts)
  singular = !texts.is_a?(Array)
  texts = [texts] if singular

  result = []
  texts.each do |text|
    # tokenize
    tokens, start_offsets, end_offsets = @tokenizer.text_to_ids_with_offsets(text, nil, 100) # unk token
    tokens.unshift(101) # cls token
    tokens << 102 # sep token

    # infer
    input = {
      input_ids: [tokens],
      attention_mask: [[1] * tokens.size],
      token_type_ids: [[0] * tokens.size]
    }
    output = @model.predict(input)

    # transform
    entities = output["output_0"][0]
    score =
      entities.map do |e|
        values = e.map { |v| Math.exp(v) }
        sum = values.sum
        values.map { |v| v / sum }
      end

    labels_idx = score.map { |s| s.each_with_index.max[1] }
    labels = ["O", "B-MISC", "I-MISC", "B-PER", "I-PER", "B-ORG", "I-ORG", "B-LOC", "I-LOC"]

    entities = []
    filtered_label_idx = labels_idx.map.with_index.reject { |v, i| v == 0 }
    filtered_label_idx.each do |label_idx, idx|
      entities << {
        score: score[idx][label_idx],
        entity: labels[label_idx],
        index: idx
      }
    end

    result << group_entities(entities, text, start_offsets, end_offsets)
  end

  singular ? result.first : result
end