Exploration and production document content and metadata scanner

Patent metadata

Field	Value
Publication number	US-12437570-B2
Application number	US-202218260526-A
Country	US
Kind code	B2
Filing date	Jan 7, 2022
Priority date	Jan 8, 2021
Publication date	Oct 7, 2025
Grant date	Oct 7, 2025

Abstract

A method involves extracting, from a file comprising an unstructured oilfield document, terms, calculating term frequency inverse document frequency (TF-IDF) of the terms to generate an input vector, execute a document content classification model on the input vector to generate a document content classification of unstructured oilfield document, and extract table information from a table in the unstructured oilfield document. The method further involves storing, with the file in storage, the document content classification and the table information.

First claim

What is claimed is:

1. A method comprising: obtaining, for a plurality of oilfield document content classes, a training set comprising a plurality of documents; calculating an inverse document frequency from the plurality of documents in the training set; calculating term frequency inverse document frequency (TF-IDF) of terms in the training data set to generate a plurality of TF-IDF vector results related to a plurality of document content classes; training the document content type classification model using the plurality of TF-IDF vector results; extracting, from a file comprising an unstructured oilfield document, a plurality of terms; calculating TF-IDF of the plurality of terms to generate an input vector; executing a document content classification model on the input vector to generate a document content classification of unstructured oilfield document; extracting table information from a table in the unstructured oilfield document; and storing, with the file in storage, the document content classification and the table information.

2. The method of claim 1 , wherein the document content classification comprises:
a plurality of document content classes each associated with a corresponding probability of the unstructured oilfield document being in the document content class.

3. The method of claim 1 , wherein extracting table information comprises:
detecting a table in the unstructured oilfield document;
generating a bounding box around the table;
detecting a plurality of rows and a plurality of columns of the table using the bounding box;
extracting contents from the plurality of rows and the plurality of columns;
interrelating the contents in the plurality of rows to obtain related contents; and
storing the related contents in a comma separated value file.

4. The method of claim 3 , further comprising:
obtaining, from a table control file, a table parameter of the table, wherein the table parameter specifies whether the table comprises a plurality of vertical lines,
detecting the plurality of vertical lines in the table based on the table parameter; and
wherein detecting the plurality of columns is performed using the plurality of vertical lines.

5. The method of claim 3 , further comprising:
obtaining, from a table control file, a table parameter of the table, wherein the table parameter specifies whether the table comprises a plurality of horizontal lines,
detecting the plurality of horizontal lines in the table based on the table parameter; and
wherein detecting the plurality of columns is performed using the plurality of horizontal lines.

6. The method of claim 1 , further comprising:
obtaining a control file comprising:
a model specification of the document type classification model, and
a data extraction control file path specifying a location to store the document content classification and the table information.

7. The method of claim 1 , further comprising:
extracting file metadata of the file; and
cataloging the unstructured oilfield document using the file metadata.

8. A system comprising: memory; and a processor for executing computer readable code configured to perform operations comprising: obtaining, for a plurality of oilfield document content classes, a training set comprising a plurality of documents; calculating an inverse document frequency from the plurality of documents in the training set; calculating term frequency inverse document frequency (TF-IDF) of terms in the training data set to generate a plurality of TF-IDF vector results related to a plurality of document content classes; training the document content type classification model using the plurality of TF-IDF vector results; extracting, from a file comprising an unstructured oilfield document, a plurality of terms, calculating TF-IDF of the plurality of terms to generate an input vector, executing a document content classification model on the input vector to generate a document content classification of unstructured oilfield document, extracting table information from a table in the unstructured oilfield document, and storing, with the file in storage, the document content classification and the table information.

9. The system of claim 8 , wherein the document content classification comprises:
a plurality of document content classes each associated with a corresponding probability of the unstructured oilfield document being in the document content class.

10. The system of claim 8 , wherein extracting table information comprises:
detecting a table in the unstructured oilfield document;
generating a bounding box around the table;
detecting a plurality of rows and a plurality of columns of the table using the bounding box;
extracting contents from the plurality of rows and the plurality of columns;
interrelating the contents in the plurality of rows to obtain related contents; and
storing the related contents in a comma separated value file.

11. The system of claim 10 , the operations further comprising:
obtaining, from a table control file, a table parameter of the table, wherein the table parameter specifies whether the table comprises a plurality of vertical lines,
detecting the plurality of vertical lines in the table based on the table parameter; and
wherein detecting the plurality of columns is performed using the plurality of vertical lines.

12. The system of claim 10 , the operations further comprising:
obtaining, from a table control file, a table parameter of the table, wherein the table parameter specifies whether the table comprises a plurality of horizontal lines,
detecting the plurality of horizontal lines in the table based on the table parameter; and
wherein detecting the plurality of columns is performed using the plurality of horizontal lines.

13. The system of claim 8 , the operations further comprising:
obtaining a control file comprising:
a model specification of the document type classification model, and
a data extraction control file path specifying a location to store the document content classification and the table information.

14. The system of claim 8 , the operations further comprising:
extracting file metadata of the file; and
cataloging the unstructured oilfield document using the file metadata.

15. A non-transitory computer readable medium comprising instructions that, when executed by a computer processor, perform operations comprising: obtaining, for a plurality of oilfield document content classes, a training set comprising a plurality of documents; calculating an inverse document frequency from the plurality of documents in the training set; calculating term frequency inverse document frequency (TF-IDF) of terms in the training data set to generate a plurality of TF-IDF vector results related to a plurality of document content classes; training the document content type classification model using the plurality of TF-IDF vector results; extracting, from a file comprising an unstructured oilfield document, a plurality of terms; calculating TF-IDF of the plurality of terms to generate an input vector; executing a document content classification model on the input vector to generate a document content classification of unstructured oilfield document; extracting table information from a table in the unstructured oilfield document; and storing, with the file in storage, the document content classification and the table information.

16. The non-transitory computer readable medium of claim 15 , wherein the document content classification comprises:
a plurality of document content classes each associated with a corresponding p

Assignees

• Schlumberger Technology Corp

Inventors

• Gupta Rishabh
• Patel Swapnil
• Sinha Udit

Classifications

• G06V2201/10

  Recognition assisted with metadata · CPC title

• G06V30/414

  Extracting the geometrical structure, e.g. layout tree; Block segmentation, e.g. bounding boxes for graphics or text · CPC title

• G06V30/42

  based on the type of document · CPC title

• G06V30/412 Primary

  Layout analysis of documents structured with printed lines or input boxes, e.g. business forms or tables · CPC title

• G06V30/19147

  Obtaining sets of training patterns; Bootstrap methods, e.g. bagging or boosting · CPC title