System and method for generating legal document and computer program product thereof

Abstract

A legal document generation system, method, and computer program product include receiving voice messages and storing them as audio files; recognizing the audio files to convert them into text content; segmenting and structuring the text content according to preset keywords to generate structured paragraph data; receiving document data; performing character recognition and field extraction on the document data to generate structured document data; performing named entity recognition and semantic fusion analysis on the segmented text content in the structured paragraph data and the structured document data to generate structured summary data; and filling the structured summary data into a preset template to generate a legal document containing complete paragraphs, legal citations, and signature fields.

Claims

1. A legal document generation system, comprising: A voice recording module is used to receive a voice message and store it as an audio file. The system comprises the following modules: a speech recognition module for converting audio files from the speech acquisition module into text; a content segmentation module for segmenting and structuring the text based on preset keywords to generate structured segment data containing segment type, segment content, and semantic tags; a document acquisition module for receiving document data; a character recognition module for recognizing characters and extracting fields from the document data in the document acquisition module to generate structured document data containing field names, field content, and location coordinates; a summary generation module for performing named entity recognition and semantic fusion analysis on the segmented text content in the structured segment data of the content segmentation module and the structured document data of the character recognition module to generate structured summary data containing summary segment type, summary segment content, semantic classification tags, and named entities; and a document generation module for filling the structured summary data into a preset template to generate a legal document.

2. The legal document generation system as described in claim 1, wherein, The abstract generation module performs named entity recognition on the segmented text content in the structured paragraph data and the structured document data to identify named entities, and performs semantic analysis on the semantic fusion result to mark the purpose of each segment.

3. The legal document generation system as described in claim 1, wherein, The speech recognition module further stores the audio file as structured data containing the text content and the corresponding timestamp. The content segmentation module further segments and marks the text content of the structured data according to the preset keywords and the timestamp to generate the structured paragraph data.

4. The legal document generation system as described in claim 1, wherein, The document collection module is further used to perform format verification and integrity checks on the document data. The format verification includes verifying whether the format of the document data is a format supported by the system, and the integrity check includes checking whether the document data is missing pages or damaged.

5. A method for generating legal documents, comprising: Receive a voice message and save it as an audio file; The audio file is converted into text content through speech recognition; the text content is segmented and structured according to preset keyword classification to generate structured paragraph data containing paragraph type, paragraph content, and semantic tags; a document is received; the document is identified and its fields are extracted to generate structured document data containing field names, field content, and location coordinates; the segmented text content in the structured paragraph data and the structured document data are analyzed using named entity recognition and semantic fusion to generate structured summary data containing summary paragraph type, summary paragraph content, semantic classification tags, and named entities; and the structured summary data is filled into a preset template to generate a legal document.

6. The legal document generation method as described in claim 5, wherein, The steps of performing named entity recognition and semantic fusion analysis further include: performing named entity recognition on the segmented text content in the structured paragraph data and the structured document data to identify the named entity, and performing semantic analysis on the semantic fusion result to mark the purpose of each segment.

7. The legal document generation method as described in claim 5, wherein, The step of converting the text content further includes: storing it as structured data containing the text content and the corresponding timestamp.

8. The legal document generation method as described in claim 7, wherein, The steps of segmenting and structuring the data further include: segmenting and marking the text content of the structured data according to the preset keywords and the timestamp to generate the structured paragraph data.

9. The legal document generation method as described in claim 5, wherein, The step of receiving file data further includes: performing format verification and integrity checks on the file data, wherein the format verification includes verifying whether the format of the file data is a format supported by the system, and the integrity check includes checking whether the file data is missing pages or damaged.

10. A computer program product that loads a program onto a computer to perform the legal document generation method described in any one of claims 6 to 9.

Images