Recognition of Chinese Legal Elements Based on Transfer Learning and Semantic Relevance

Algorithm 1

Data preprocessing.

Input: a list of each sentence and corresponding label in the legal document doc, ;

Output: Each sentence of the legal instrument and all combinations of elements, and a 0, 1 label. If the label is 0, then the sentence does not contain the element; if the label is 1, the sentence contains the element.

1.

2. For sen in doc do

3. For i =1 to 20

4.  Algorithm 1 Data preprocessingif lab_i in data[sen]: l =1

5. Else: l =0

6. T.attend({l, (sen, lab_i)})

7. End for

8. End for

9. Return T