Unlock instant, AI-driven research and patent intelligence for your innovation.

Semantic coding method and apparatus for network resources

A technology of network resources and semantic coding, applied in semantic analysis, network data indexing, network data retrieval, etc., can solve the problems of inconsistency in semantic description, difficulty in guaranteeing the accuracy of web words, and inapplicability to user behavior data, etc.

Inactive Publication Date: 2016-04-13
ALIBABA (CHINA) CO LTD
View PDF8 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] Since existing technologies such as Word2vec use a single word as the basic processing unit, but the expression of a sentence or phrase is obtained by combining the semantic expressions of words, the accuracy of the webpage words obtained in this way is difficult to guarantee
Moreover, the word vector definition method in the prior art requires a large amount of webpage text data as a training corpus, and the obtained result is also a semantic representation in the general sense, which is different from the semantic description required in vertical fields such as video
Existing methods are not suitable for processing user behavior data such as query clicks

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Semantic coding method and apparatus for network resources
  • Semantic coding method and apparatus for network resources
  • Semantic coding method and apparatus for network resources

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0125] figure 1 It is a flowchart of a method for semantic encoding of network resources according to an embodiment of the present invention. Such as figure 1 As shown, the network resources may include multimedia resources that can be accessed through the Internet and user behavior data generated by users accessing the multimedia resources, such as web pages such as videos, audios, and pictures that the user can access through the Internet. The user behavior data may include input data and click data, and the input data may be, for example, search terms input by the user into a search engine. The click data may be, for example, counting the number of clicks obtained by the user by clicking on certain audio, video, and pictures.

[0126] The semantic coding method of the network resource can mainly include:

[0127] Step 101: Determine the degree of association between every two network resources in the area to be processed according to the multimedia resources, input data and c...

Embodiment 2

[0137] image 3 It is a flowchart of a method for semantic encoding of network resources according to another embodiment of the present invention. image 3 Win the mark and figure 1 , figure 2 The same steps have the same functions. For brevity, detailed descriptions of these steps are omitted.

[0138] Such as image 3 As shown, the difference between this embodiment and the previous embodiment is that step 101 may include:

[0139] Step 201: According to the multimedia resource, the input data, and the click data, establish an initial association relationship diagram of each of the network resources in the area to be processed.

[0140] Step 202: Perform an iterative operation according to the initial association relationship graph, the multimedia resource, the input data, and the click data, and adjust the initial association relationship graph according to the result of the iterative operation.

[0141] Step 203: Determine the degree of association between the multimedia resource...

Embodiment 3

[0218] Picture 11 It is a structural block diagram of an apparatus for semantic encoding of network resources according to an embodiment of the present invention. Wherein, the network resources include multimedia resources that can be accessed through the Internet and user behavior data generated by users accessing the multimedia resources, for example, web pages such as videos, audios, and pictures that the user can access through the Internet. The user behavior data includes input data and click data, and the input data may be, for example, search terms input by the user into a search engine. The click data may be, for example, counting the number of clicks obtained by the user by clicking on certain audio, video, and pictures.

[0219] Such as Picture 11 As shown, the semantic encoding device for network resources may mainly include:

[0220] The degree of association determining module 11 is configured to determine the degree of association of every two of the network resour...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention relates to a semantic coding method and apparatus for network resources. The network resources include multimedia resources accessible through the internet and user behavior data generated by accessing to the multimedia resources by users, and the user behavior data include input data and click data. The method comprises: according to the multimedia resources, the input data and the click data, determining a degree of correlation of every two network resources in a to-be-processed region; and according to the degree of correlation of every two network resources, performing semantic coding on the multimedia resources and / or the input data, wherein a semantic coding result is that the network resources are represented with vectors. According to an embodiment of the invention, the obtained degree of correlation can accurately reflect user behaviors, so that the network resources can be accurately subjected to semantic coding according to the degree of correlation of every two network resources to obtain semantic vectors of the network resources.

Description

Technical field [0001] The present invention relates to the Internet field, in particular to a method and device for semantic encoding of network resources. Background technique [0002] When searching on a web page using search terms, it is not easy to retrieve a word related to the search term from a large number of webpage terms, and there may be problems such as low correlation between the searched webpage term and the expected search term. [0003] At present, the correlation between search terms and webpage words can be quickly found through accurate word vectors. Word2vec is a tool that can convert individual words into vector form. Specifically, Word2vec infers the semantic relationship between words by mining the positional relationship between words in a large number of web pages, including location adjacent, similar location, co-occurrence, etc., and expresses this semantic relationship with vectors . [0004] Since the prior art such as Word2vec uses a single word as a...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): G06F17/30G06F17/27
CPCG06F16/951G06F40/30
Inventor 邹敏齐志兵尹玉宗姚键潘柏宇王冀
Owner ALIBABA (CHINA) CO LTD