# US 2013/0158981 A1 · Linking newsworthy events to published content

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

- Number: US 2013/0158981 A1
- Document ID: US20130158981A1
- Status: Published application
- Kind: A1
- Application number: 13/331,215
- Filed: 2011-12-20
- Published: 2013-06-20
- Earliest claimed: 2011-12-20
- Assignees: Yahoo! Inc.
- Inventors: Hakan Ceylan; Ergin Elmacioglu; Meryem Pinar Donmez
- Web page: https://patentgrep.com/patent/US20130158981A1

## CPC

- G06F 40/134: Hyperlinking
- G06F 16/958: Organisation or management of web site content, e.g. publishing, maintaining pages or automatic linking
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

Methods, systems, and computer programs are presented for linking newsworthy events in a document to published content. One method includes an operation for receiving features by a classifier that is operable to determine a probability of the availability of news for a sentence. When the features are found in the sentence, the probability of the availability of news for the sentence increases, where the sentence includes one or more noun phrases and ends in a full stop. The classifier determines which sentences in a document are candidate sentences for being linked to news articles, and for each candidate sentence, the method includes an operation for finding an associated news article when there is an associated news article exceeding a relevance threshold. Further, the method includes operations for adding links in the document to the found associated news articles, and for displaying the document with the added links.

## Claims

### Claim 1

A method for providing internet content, the method comprising:
receiving a plurality of features by a classifier that is operable to determine a probability of an availability of news for a sentence, wherein at least one of the features when found in the sentence increases a probability of the availability of news for the sentence, the sentence including one or more noun phrases and ending in a full stop;
determining by the classifier which sentences in a document are candidate sentences for having available news;
for each candidate sentence, finding an associated news article when the associated news article exceeds a threshold of relevance to the candidate sentence;
adding links in the document to the found associated news articles; and
displaying the document with the added links, wherein at least one operation of the method is executed through a processor.

### Claim 2 (depends on 1)

The method as recited in claim 1, further including:
receiving judgments for a plurality of training sentences; and
training the classifier utilizing machine learning with the plurality of training sentences, the judgments, and the received plurality of features.

### Claim 3 (depends on 2)

The method as recited in claim 2, wherein training the classifier further includes:
identifying a value for each feature, when found in the sentence, towards determining the availability of news for the sentence.

### Claim 4 (depends on 1)

The method as recited in claim 1, wherein determining which sentences are candidate sentences further includes:
appraising each sentence with the classifier based on the sentence and the plurality of features; and
determining if each sentence is a candidate sentence based on the appraising.

### Claim 5 (depends on 1)

The method as recited in claim 1, wherein finding the associated news article further includes:
for each candidate sentence, performing a search on the candidate sentence to find potential news articles;
evaluating each potential news article against the candidate sentence to obtain a score for each potential news article based on a probability that the potential news article has information about the candidate sentence; and
selecting a best potential news article with a highest score when the score of the best potential news article exceeds a threshold score.

### Claim 6 (depends on 5)

The method as recited in claim 5, wherein evaluating each potential news article further includes:
comparing words in the sentence to words in a title and abstract of the potential news article.

### Claim 7 (depends on 5)

The method as recited in claim 5, wherein evaluating each potential news article further includes:
comparing words in the sentence to words in a title, abstract, and body of the potential news article.

### Claim 8 (depends on 5)

The method as recited in claim 5, wherein evaluating each potential news article further includes:
utilizing a relevance model to evaluate the potential news article against the sentence.

### Claim 9 (depends on 5)

The method as recited in claim 5, wherein the search is limited to news occurring within a time period having a predetermined time length of time and ending in a current time.

### Claim 10 (depends on 5)

The method as recited in claim 5, wherein adding links in the document further includes:
adding a link in each candidate sentence with the associated news article to the associated news article, the link being added to a plurality of words in the candidate sentence, the plurality of words including a verb and a noun phrase.

### Claim 11

A method for providing internet content, the method comprising:
training a classifier utilizing a plurality of features and a plurality of training sentences with associated judgments, wherein the classifier is operable to determine a probability of an availability of news for a sentence, wherein at least one of the features when found in the sentence increases a probability of the availability of news for the sentence;
determining by the classifier which sentences in a document are candidate sentences for having available news;
for each candidate sentence, finding an associated news article when the associated news article exceeds a threshold of relevance to the candidate sentence;
adding links in the document to the found associated news articles; and
displaying the document with the added links, wherein at least one operation of the method is executed through a processor.

### Claim 12 (depends on 11)

The method as recited in claim 11, wherein finding the associated news article further includes:
eliminating news as possible associated news articles when the news are duplicates of the document.

### Claim 13 (depends on 11)

The method as recited in claim 11, wherein the associated news article provides news associated with a corresponding sentence.

### Claim 14 (depends on 11)

The method as recited in claim 11, wherein operations of the method are performed by a computer program when executed by one or more processors, the computer program being embedded in a non-transitory computer-readable storage medium.

### Claim 15 (depends on 11)

The method as recited in claim 11, wherein each sentence in the document starts with a capital letter and includes one or more verbs.

### Claim 16

A computer program embedded in a non-transitory computer-readable storage medium, when executed by one or more processors, for providing internet content, the computer program comprising:
program instructions for receiving a plurality of features by a classifier that is operable to determine a probability of an availability of news for a sentence, wherein at least one of the features when found in the sentence increases a probability of the availability of news for the sentence, the sentence including one or more noun phrases and ending in a full stop;
program instructions for determining by the classifier which sentences in a document are candidate sentences for having available news;
program instructions for finding, for each candidate sentence, an associated news article when the associated news article exceeds a threshold of relevance to the candidate sentence;
program instructions for adding links in the document to the found associated news articles; and
program instructions for displaying the document with the added links.

### Claim 17 (depends on 16)

The computer program as recited in claim 16, wherein determining which sentences are candidate sentences further includes:
program instructions for appraising each sentence with the classifier based on the sentence and the plurality of features; and
program instructions for determining if each sentence is a candidate sentence based on the appraising.

### Claim 18 (depends on 16)

The computer program as recited in claim 16, wherein finding the associated news article further includes:
program instructions for performing, for each candidate sentence, a search on the candidate sentence to find potential news articles;
program instructions for evaluating each potential news article against the candidate sentence; and
program instructions for selecting a best potential news article based on the evaluation to be the associated news article when a score of the best potential news article exceeds a threshold score.

### Claim 19 (depends on 18)

The computer program as recited in claim 18, wherein the features are selected from a group consisting of a date, a verb in the past tense, a proper name in the sentence, a name of a celebrity, a name of a media outlet, a name of a crime, or a title.

### Claim 20 (depends on 18)

The computer program as recited in claim 18, wherein adding links in the document further includes:
program instructions for adding a link in each candidate sentence with the associated news article to the associated news article, the link being added to a plurality of words in the candidate sentence, the plurality of words including a verb.
