Method, server, browser, and system for recommending text information

Patent metadata

Field	Value
Publication number	US-2018285331-A1
Application number	US-201816002855-A
Country	US
Kind code	A1
Filing date	Jun 7, 2018
Priority date	Jun 6, 2013
Publication date	Oct 4, 2018
Grant date	—

Abstract

A method for recommending text information, a server, a browser, and a system thereof are provided. The method includes: a keyword is obtained by a server from a text index pool in response to a visiting request for a webpage from a browser; a recommended text information list is obtained from the text index pool according to the corresponding keyword; the recommended text information list is embedded into the webpage and the webpage is returned to the browser for displaying to a user; and the user may click the keyword to obtain the recommended text information list as required, thereby the keyword is set as the connection between the user and the webpage text information.

First claim

1 - 20 . (canceled)

21 . A method for recommending text information, comprising:
extracting keywords from a plurality of webpages, including extracting a first keyword from at least one of the plurality of webpages, wherein each of the plurality of webpages corresponding to at least one of the keywords; obtaining multiple webpages corresponding to the first keyword; sorting the multiple webpages to obtain a sorting result; taking a predetermined number of the multiple webpages as recommended webpages corresponding to the first keyword based on the sorting result; creating a text index pool by taking Uniform Resource Locators (URLs) of the multiple webpages as Keys of the text index pool and taking the first keyword as a Value of the text index pool; creating a keyword index pool by taking the first keyword as a Key of the keyword index pool and taking the URLs of the recommended webpages as Values of the keyword index pool; obtaining, in response to a visiting request for a webpage at a URL, a target keyword corresponding to the URL of the webpage; obtaining a recommended text list corresponding to the target keyword from the keyword index pool; and embedding the recommended text list into the webpage.

22 . The method according to claim 21 , wherein extracting the first keyword from at least one of the plurality of webpages comprises:
extracting a main text from a webpage of the plurality of webpages; performing word segmentation on the main text to obtain candidate keywords; counting a word frequency of each of the candidate keywords and a distribution parameter of each of the candidate keywords; calculating a weight of each of the candidate keywords based on the word frequency and the distribution parameter; and setting a candidate keyword of which the weight is greater than a predetermined threshold as the first keyword.

23 . The method according to claim 21 , wherein sorting the multiple webpages and taking the predetermined number of the multiple webpages as recommended webpages comprises:
calculating a text weight of each of the multiple webpages by taking a number of historical hits and an updating time of each of the multiple webpages as parameters of each of the multiple webpages; and setting the predetermined number of multiple webpages as the recommended webpages based on text weights of the multiple webpages.

24 . The method according to claim 21 , further comprising:
obtaining the recommended text list corresponding to the target keyword by querying the keyword index pool using the target keyword as a Key, wherein the recommended text list includes recommended webpages stored in the keyword index pool and correspond to the target keyword.

25 . The method according to claim 21 , wherein obtaining the target keyword corresponding to the URL of the webpage comprises:
determining whether the text index pool includes the URL of the webpage by querying the text index pool using the URL of the webpage as a Key; obtaining the target keyword corresponding to the URL of the webpage from the text index pool if the text index pool includes the URL of the webpage; and obtaining the target keyword corresponding to the URL of the webpage by performing word segmentation and selecting a segmented word based on appearance frequency and distribution manner of the segmented word in the webpage if the text index pool does not include the URL of the webpage.

26 . The method according to claim 21 , wherein embedding the recommended text list into the webpage comprises:
linking the recommended text list with the target keyword; and marking the target keyword at a corresponding position of the webpage.

27 . The method according to claim 26 , further comprising:
in response to the target keyword being selected at the corresponding position of the webpage in a browser, sending the recommended text list embedded in the webpage to the browser for displaying.

28 . A server for recommending text information, comprising a memory; and a processor coupled to the memory, and wherein the processor is configured to:
extract keywords from a plurality of webpages, including extracting a first keyword from at least one of the plurality of webpages, wherein each of the plurality of webpages corresponding to at least one of the keywords; obtain multiple webpages corresponding to the first keyword; sort the multiple webpages to obtain a sorting result; take a predetermined number of the multiple webpages as recommended webpages corresponding to the first keyword based on the sorting result; create a text index pool by taking Uniform Resource Locators (URLs) of the multiple webpages as Keys of the text index pool and taking the first keyword as a Value of the text index pool; create a keyword index pool by taking the first keyword as a Key of the keyword index pool and taking the URLs of the recommended webpages as Values of the keyword index pool; obtain, in response to a visiting request for a webpage at a URL, a target keyword corresponding to the URL of the webpage; obtain a recommended text list corresponding to the target keyword from the keyword index pool; and embed the recommended text list into the webpage.

29 . The server according to claim 28 , wherein the processor is further configured to:
extract a main text from a webpage of the plurality of webpages; perform word segmentation on the main text to obtain candidate keywords; count a word frequency of each of the candidate keywords and a distribution parameter of each of the candidate keywords; calculate a weight of each of the candidate keywords based on the word frequency and the distribution parameter; and set a candidate keyword of which the weight is greater than a predetermined threshold as the first keyword.

30 . The server according to claim 28 , wherein the processor is further configured to:
calculate a text weight of each of the multiple webpages by taking a number of historical hits and an updating time of each of the multiple webpages as parameters of each of the multiple webpages; and set the predetermined number of multiple webpages as the recommended webpages based on text weights of the multiple webpages.

31 . The server according to claim 28 , wherein the processor is further configured to:
obtain the recommended text list corresponding to the target keyword by querying the keyword index pool using the target keyword as a Key, wherein the recommended text list includes recommended webpages stored in the keyword index pool and correspond to the target keyword.

32 . The server according to claim 28 , wherein the processor is further configured to:
determine whether the text index pool includes the URL of the webpage by querying the text index pool using the URL of the webpage as a Key; obtain the target keyword corresponding to the URL of the webpage from the text index pool if the text index pool includes the URL of the webpage; and obtain the target keyword corresponding to the URL of the webpage by performing word segmentation and selecting a segmented word based on appearance frequency and distribution manner of the segmented word in the webpage if the text index pool does not include the URL of the webpage.

33 . The server according to claim 28 , wherein the processor is further configured to:
link the recommended text list with the target keyword; and mark the target keyword at a corresponding position of the webpage.

34 . The server according to claim 28 , wherein the processor is further configured to: in response to the target keyword being selected at the corresponding positi

Assignees

• Tencent Tech Shenzhen Co Ltd

Inventors

• Cai Bing

Classifications

• G06F16/951 Primary

  Indexing; Web crawling techniques · CPC title

• G06F40/12

  Use of codes for handling textual entities · CPC title

• G06F40/14

  Tree-structured documents (parsing G06F40/205; validation G06F40/226) · CPC title

• G06F40/169

  Annotation, e.g. comment data or footnotes · CPC title

• G06Q30/02

  Marketing; Price estimation or determination; Fundraising · CPC title