Relationship analysis and mapping for interrelated multi-layered datasets

Patent metadata

Field	Value
Publication number	US-11874850-B2
Application number	US-202217723460-A
Country	US
Kind code	B2
Filing date	Apr 19, 2022
Priority date	Dec 7, 2017
Publication date	Jan 16, 2024
Grant date	Jan 16, 2024

Abstract

A system stores original datasets in a datastore. The system generates first derivative datasets from the original datasets, and generates second derivative datasets from at least the first derivative datasets. The system determines relationships among the original datasets, the first derivative datasets, and the second derivative datasets, based on an analytical relationship between two datasets, a similarity relationship between two datasets, a modification relationship between two datasets, and a user-interaction relationship between two datasets. Then, the system generates a node map including at least part of the original datasets, the first derivative datasets, and the second derivative datasets as a node, and at least part of the determined analytical, similarity, modification, and user-interaction relationships between two nodes as a link.

First claim

The invention claimed is:

1. A system comprising:
one or more hardware processors; and
a memory storing instructions that, when executed by the one or more hardware processors, cause the system to perform:
translating an original dataset of an original language into a first translated dataset of another language;
responsive to determining that a description of the first translated dataset fails to match a corresponding description of the original dataset, generating a mapping of keywords in the original language to translated keywords of the translated language;
translating the original dataset to a second translated dataset based on the mapping;
generating and displaying a focused node map in a main field of a graphical user interface (GUI), wherein the focused node map comprises a selected node of the second translated dataset satisfying a filtering criterion and one or more nodes linked with the selected node; and
displaying searched or selected data of the second translated dataset in an auxiliary field of the GUI, the auxiliary field being presented at a side of the main field, the auxiliary field further comprising any of:
a content field that provides a portion of the focused node map indicating a determined relationship of the second translated dataset with another dataset among the original dataset, the first translated dataset, or a different dataset; and
a metadata field that presents metadata of the second translated dataset including a data source and the other dataset.

2. The system of claim 1 , wherein the metadata further comprises respective paths and sources of the dataset and the another dataset among the original dataset, the first translated dataset, or the different dataset.

3. The system of claim 1 , wherein the focused node map comprises the nodes connected by links, and respective lengths of the links are indicative of proximity degrees among datasets represented by the nodes disposed at ends of each of the links.

4. The system of claim 1 , wherein the translating of the original dataset to the second translated dataset comprises:
performing a speech recognition process on the original dataset; and
performing a term-frequency-inverse-document frequency (TF-IDF) analysis on an audio data field of the original dataset in response to the performing of the speech recognition process.

5. The system of claim 1 , wherein the selected node in the focused node map is visualized with emphasis based on a proximity of a relationship between the selected node and at least one of the linked nodes.

6. The system of claim 1 , wherein the auxiliary field comprises a first auxiliary field; and the instructions further cause the one or more processors to perform:
opening a second auxiliary field that indicates the filtering criterion; and
decreasing a size of the main field in response to the opening of the second auxiliary field.

7. The system of claim 1 , wherein the focused node map is populated in response to a selection of the selected node.

8. The system of claim 1 , wherein the original dataset comprises a video file; and the instructions further cause the one or more processors to perform:
extracting parameters by performing image analysis of the video file; and
determining the relationship between the second translated dataset with the other dataset based on a comparison between the extracted parameters and corresponding parameters of the other dataset.

9. The system of claim 1 , wherein the instructions further cause the one or more processors to perform:
populating, in the main field, a downstream dataset generated from analysis of the second translated dataset and successive downstream datasets, each of which is generated from analysis of a preceding downstream dataset.

10. The system of claim 1 , wherein the instructions further cause the one or more processors to perform:
populating, in the main field, an upstream dataset from which the original dataset was generated, and successive upstream datasets linked to one another via analysis of an immediate preceding upstream dataset.

11. A method performed on a computer system having one or more hardware processors programmed with computer program instructions that, when executed by the one or more hardware processors, cause the computer system to perform the method, the method comprising:
translating an original dataset of an original language into a first translated dataset of another language;
responsive to determining that a description of the first translated dataset fails to match a corresponding description of the original dataset, generating a mapping of keywords in the original language to translated keywords of the translated language;
translating the original dataset to a second translated dataset based on the mapping;
generating and displaying a focused node map in a main field of a graphical user interface (GUI), wherein the focused node map comprises a selected node of the second translated dataset satisfying a filtering criterion and one or more nodes linked with the selected node; and
displaying searched or selected data of the second translated dataset in an auxiliary field of the GUI, the auxiliary field being presented at a side of the main field, the auxiliary field further comprising any of:
a content field that provides a portion of the focused node map indicating a determined relationship of the second translated dataset with another dataset among the original dataset, the first translated dataset, or a different dataset; and
a metadata field that presents metadata of the second translated dataset including a data source and the other dataset.

12. The method of claim 11 , wherein the metadata further comprises respective paths and sources of the dataset and the another dataset among the original dataset, the first translated dataset, or the different dataset.

13. The method of claim 11 , wherein the focused node map comprises the nodes connected by links, and respective lengths of the links are indicative of proximity degrees among datasets represented by the nodes disposed at ends of each of the links.

14. The method of claim 11 , wherein the translating of the original dataset to the second translated dataset comprises:
performing a speech recognition process on the original dataset; and
performing a term-frequency-inverse-document frequency (TF-IDF) analysis on an audio data field of the original dataset in response to the performing of the speech recognition process.

15. The method of claim 11 , wherein the selected node in the focused node map is visualized with emphasis based on a proximity of a relationship between the selected node and at least one of the linked nodes.

16. The method of claim 11 , wherein the auxiliary field comprises a first auxiliary field; and the method further comprises:
opening a second auxiliary field that indicates the filtering criterion; and
decreasing a size of the main field in response to the opening of the second auxiliary field.

17. The method of claim 11 , wherein the focused node map is populated in response to a selection of the selected node.

18. The method of claim 11 , wherein the original dataset comprises a video file; and the method further comprises:
extracting parameters by performing image analysis of the video file; and
determining the relationship between the second translated dataset with the other dataset based on a comparison between the extracted parameters and corresponding parameters of the other dataset.

19. The method of claim 11 , further compri

Assignees

• Palantir Technologies Inc

Inventors

• Prakash Arjun
• Moore Becky
• Pilch Jakub

Classifications

• G06F16/26 Primary

  Visual data mining; Browsing structured data · CPC title

• G06F16/2264

  Multidimensional index structures · CPC title

• G06F16/288

  Entity relationship models · CPC title

• G06F16/904 Primary

  Browsing; Visualisation therefor (for navigating the web G06F16/954; browsing optimisation for the web G06F16/957) · CPC title