# US 2020/0004808 A1 · Document editing models and management

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2020/0004808 A1
- Document ID: US20200004808A1
- Status: Published application
- Kind: A1
- Application number: 16/020,994
- Filed: 2018-06-27
- Published: 2020-01-02
- Earliest claimed: 2018-06-27
- Assignees: Microsoft Technology Licensing, LLC
- Applicants: Microsoft Technology Licensing, LLC
- Inventors: Zhenyu YAO; Sharon Hang LI; Maninder jit Singh DHALIWAL; Karvell Ka Yiu LI; Yimeng LI
- Web page: https://patentgrep.com/patent/US20200004808A1

## CPC

- G06F 40/166: Editing, e.g. inserting or deleting
- G06F 3/0482: Interaction with lists of selectable items, e.g. menus
- G06F 40/106: Display of layout of documents; Previewing
- G06F 40/232: Orthographic correction, e.g. spell checking or vowelisation
- G06N 3/0464: Convolutional networks [CNN, ConvNet]
- G06N 3/0895: Weakly supervised learning, e.g. semi-supervised or self-supervised learning
- G06N 3/09: Supervised learning
- G06N 3/091: Active learning
- G06N 20/00: Machine learning
- G06N 3/045: Combinations of networks
- G06N 5/01: Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
- G06N 7/01: Probabilistic graphical models, e.g. probabilistic networks
- G06N 20/10: using kernel methods, e.g. support vector machines [SVM]
- G06N 20/20: Ensemble learning
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

A mechanism and model by which to manage inadvertent edits of electronic content. The model determines the likelihood of whether a modification made to a document is intentional or was inadvertent and assist users in identifying these edits. The proposed devices and method can significantly improve workflow efficiency and allow users to feel more comfortable in the development and use of their electronic content, and/or reduce the possibility of other users causing accidental or unintentional errors in a document.

## Claims

### Claim 1

A device comprising:
one or more processors; and
one or more computer readable media including instructions which, when executed by the one or more processors, cause the one or more processors to:
determine that a first triggering event has occurred during access of a first electronic content, the first triggering event including a modification of a portion of the first electronic content by a first user from an initial state to a modified state;
determine, in response to the first triggering event, that the modification was inadvertent based on at least a usage history for the first user or a usage history for the first electronic content;
cause, in response to the determination that the modification was inadvertent, a first user interface element including a first option and a second option to be presented to the first user; and
revert, in response to receiving a first user input indicating a selection of the first option, the portion of the first electronic content from the modified state to the initial state.

### Claim 2 (depends on 1)

The device of claim 1, wherein the instructions further cause the one or more processors to automatically reduce access of the first user to the first electronic content from a read-write access to a read-only access in response to receiving the first user input.

### Claim 3 (depends on 1)

The device of claim 1, wherein the determination that the modification is inadvertent is further based on a determination that the modification includes a text-based typographical error.

### Claim 4 (depends on 1)

The device of claim 1, wherein the determination that the modification is inadvertent is further based on a prediction, made prior to the modification, based on at least the usage history for the first user or usage history for the first electronic content, that the accessed first electronic content will not be modified by the first user.

### Claim 5 (depends on 1)

The device of claim 1, wherein the instructions further cause the one or more processors to:
cause, in response to the determination that the modification was inadvertent, the application to present a second user interface element including a third option and a fourth option;
create, in response to receiving a second user input indicating a selection of the third option, a new file including the modified state of the first file; and
maintain, in response to receiving the second user input, the first file in its initial state.

### Claim 6

A device comprising:
one or more processors;
one or more computer readable media including instructions which, when executed by the one or more processors, cause the one or more processors to:
determine a first usage history associated with a first electronic content;
determine, by applying a first machine learning algorithm to the first usage history, that the first electronic content is associated with a low likelihood of being modified by a first user;
detect a modification of the first electronic content by the first user;
determine, upon detection of the modification of the first electronic content, that the modification was inadvertent based on the determination that the first electronic content is associated with the low likelihood of being modified by the first user;
cause, in response to the determination that the modification was inadvertent, a first user interface element including a first option and a second option to be presented to the first user; and
revert, in response to receiving a first user input indicating a selection of the first option, the modification to the first electronic content.

### Claim 7 (depends on 6)

The device of claim 6, wherein the instructions further cause the one or more processors to:
create, in response to receiving the first user input indicating a selection of the first option, a new training data sample based on at least the first usage history and indicating the modification was unintentional; and
add, at a first time, the new training data sample to a training data set;
update, at a second time after the first time and using the training data set, the first machine learning algorithm.

### Claim 8 (depends on 6)

The device of claim 6, wherein the instructions include second instructions that cause one or more of the processors to:
create, in response to receiving a second user input indicating a selection of the second option, a first training data sample based on at least the first usage history and indicating the modification was intended by the first user; and
add the first training data sample to a training data set at a first time;
update, at a second time after the first time and using the training data set, the first machine learning algorithm.

### Claim 9 (depends on 8)

The device of claim 8, wherein the second instructions further cause one or more of the processors to:
detect a reversion, by the first user, of the modification;
remove, in response to the detection of the reversion, the first training data sample from the training data set, before the second time;
create, in response to the detection of the reversion, a second training data sample based on at least the first usage history and indicating the modification was unintentional; and
add the second training data sample to the training data set before the second time.

### Claim 10 (depends on 6)

The device of claim 6, wherein the instructions further cause the one or more processors to:
detect a period of inactivity by the first user following the modification of the first electronic content; and
cause the first user interface element to be presented in response to the detection of the first period of inactivity and the determination that the modification was inadvertent.

### Claim 11

A method comprising:
determining that a first triggering event has occurred during access of a first electronic content, the first triggering event including a modification of a portion of the first electronic content by a first user from an initial state to a modified state;
determining, in response to the first triggering event, that the modification was inadvertent based on at least a usage history for the first user or a usage history for the first electronic content;
causing, in response to the determination that the modification was inadvertent, a first user interface element including a first option and a second option to be presented to the first user; and
reverting, in response to receiving a first user input indicating a selection of the first option, the portion of the first electronic content from the modified state to the initial state.

### Claim 12 (depends on 11)

The method of claim 11, wherein the method further comprises automatically reducing access of the first user to the first electronic content from a read-write access to a read-only access in response to receiving the first user input.

### Claim 13 (depends on 11)

The method of claim 11, wherein the first user interface element is presented after a delay of at least two seconds following the modification of the portion of the first electronic content.

### Claim 14 (depends on 11)

The method of claim 11, wherein the determination that the modification is inadvertent is further based on a determination that the modification includes a text-based typographical error.

### Claim 15 (depends on 11)

The method of claim 11, wherein the determination that the modification is inadvertent is further based on a prediction, made prior to the modification, based on at least the usage history for the first user or usage history for the first electronic content that the accessed first electronic content will not be modified by the first user.

### Claim 16 (depends on 11)

The method of claim 11, wherein the first user interface element includes a notification informing the first user that their most recent modification may include an inadvertent edit.

### Claim 17 (depends on 11)

The method of claim 11, further comprising:
detecting a period of inactivity by the first user following the first triggering event; and
causing the first user interface element to be presented to the first user in response to the detection of the first period of inactivity and the determination that the modification was inadvertent.

### Claim 18 (depends on 11)

The method of claim 11, further comprising:
determining that a second triggering event has occurred during access of a first file, the second triggering event including a modification of a portion of the first file by a second user from an initial state to a modified state;
determining, in response to the second triggering event, that the modification was inadvertent based on at least a usage history for the second user before the access of the first file;
causing, in response to the determination that the modification was inadvertent, a second user interface element including a third option to record the modified state as a second file to be presented to the first user;
receiving a second user input indicating a selection of the third option;
creating, in response to receiving the second user input, a new file including the modified state of the first file; and
maintaining, in response to receiving the second user input, the first file in its initial state.

### Claim 19 (depends on 11)

The method of claim 11, wherein the usage history for the first user includes data indicating an amount of time since the first electronic content was modified by the first user.

### Claim 20 (depends on 11)

The method of claim 11, wherein the usage history for the first user includes data indicating a number of occurrences that the first electronic content has been modified by the first user.
