# US 9,436,768 B2 · System and method for pushing and distributing promotion content > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 9,436,768 B2 - Document ID: US9436768B2 - Status: Granted - Kind: B2 - Application number: 14/452,518 - Filed: 2014-08-05 - Published: 2015-01-29 - Granted: 2016-09-06 - Earliest claimed: 2013-04-02 - Assignees: Tencent Technology (Shenzhen) Company Limited - Applicants: Tencent Technology (Shenzhen) Company Limited - Inventors: Pan Yang - Web page: https://patentgrep.com/patent/US9436768B2 ## CPC - G06F 16/955: using information identifiers, e.g. uniform resource locators [URL] - G06F 16/345: {Summarisation for human users} - G06F 16/353: into predefined classes - G06F 16/38: Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually - G06F 40/205: Parsing - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) ## Abstract System, methods, and computer-readable medium allow pushing content items based on traffic features. A target text is obtained and analyzed lexically. Features, including traffic feature, are extracted from the target text. Based on a pre-trained hierarchical classification model considering traffic feature, the features extracted from the target text are classified hierarchically to obtain a hierarchical classification of the target text. Based on the hierarchical classification of the target text, one or more are selected among a plurality of content items. The selected content items are obtained and pushed to a terminal. ## Claims ### Claim 1 A method for pushing content items, comprising: obtaining a target text and analyzing lexically the target text; extracting features, comprising traffic feature, from the target text; classifying hierarchically, the features extracted from the target text, based on a pre-trained hierarchical classification model considering traffic feature to obtain a hierarchical classification of the target text; selecting one or more among a plurality of content items based on the hierarchical classification of the target text; obtaining the selected content items; and pushing the obtained content items to a terminal; wherein the target text comprises a webpage text which comprises a traffic feature identifying source of the webpage text; wherein the plurality of content items comprise a plurality of promotion texts; wherein for each of the plurality of promotion texts: obtaining the promotion text and analyzing lexically the promotion text; extracting features, comprising traffic feature, from the promotion text; classifying hierarchically, the features extracted from the promotion text, based on a pre-trained hierarchical classification model considering traffic feature to obtain a hierarchical classification of the promotion text; and calculating similarity between the webpage text and the promotion text based on the hierarchical classification of the promotion text and the hierarchical classification of the webpage text, wherein the selecting one or more among the plurality of promotion texts is based on the similarity between the webpage text and each of the plurality of promotion texts. ### Claim 2 (depends on 1) The method of claim 1, in which the features extracted from the target text further comprise at least one of the following: token, keyword, topic, general keyword, expanded keyword, named entity of the target text. ### Claim 3 (depends on 1) The method of claim 1, in which the extracting features comprising traffic feature from the target text comprises: parsing Uniform Resource Locator (URL) of the webpage text; and determining traffic feature of the webpage text based on the URL. ### Claim 4 (depends on 1) The method of claim 1, in which the target text comprises a user text which comprises a traffic feature identifying source of the user text. ### Claim 5 (depends on 4) The method of claim 4, in which the user text is associated with a user identity and comprises at least one of the following: a historical search query request item associated with the user identity, a microblog tag associated with the user identity, a webpage title browsed associated with the user identity. ### Claim 6 (depends on 1) The method of claim 1, in which the target text comprises a promotion text which comprises a traffic feature identifying source of the promotion text. ### Claim 7 (depends on 1) The method of claim 1, in which the target text further comprises a user text. ### Claim 8 (depends on 7) The method of claim 7, in which the selecting one or more among the plurality of promotion texts is further based on the similarity between the webpage text and the user text. ### Claim 9 A computer-based promotion content distribution system, comprising: a lexical analysis portion configured to obtain a target text and to analyze lexically the target text; a hierarchical classification portion configured to extract features, comprising traffic feature, from the target text and to classify hierarchically the features extracted from the target text based on a pre-trained hierarchical classification model considering traffic feature to obtain a hierarchical classification of the target text; a content promotion portion configured to select one or more among a plurality of content items based on the hierarchical classification of the target text, to obtain the selected content items and to push the obtained content items to a terminal; wherein the target text comprises a webpage text which comprises a traffic feature identifying source of the webpage text; wherein the plurality of content items comprise a plurality of promotion texts; wherein the lexical analysis portion is further configured to: for each of the plurality of promotion texts, obtain the promotion text and analyze lexically the promotion text, wherein the hierarchical classification portion is further configured to: for each of the plurality of promotion texts, extract features comprising traffic feature from the promotion text and classify hierarchically the features extracted from the promotion text based on a pre-trained hierarchical classification model considering traffic feature to obtain a hierarchical classification of the promotion text, wherein the content promotion portion is configured to: for each of the plurality of promotion texts, calculate similarity between the webpage text and the promotion text based on the hierarchical classification of the promotion text and the hierarchical classification of the webpage text; and select one or more among the plurality of promotion texts based on the similarity between the webpage text and each of the plurality of promotion texts. ### Claim 10 (depends on 9) The system of claim 9, in which the hierarchical classification portion is further configured to extract from the target text at least one of the following features: token, keyword, topic, general keyword, expanded keyword, named entity of the target text. ### Claim 11 (depends on 9) The system of claim 9, in which the hierarchical classification portion is configured to: parse Uniform Resource Locator (URL) of the webpage text; and determine the traffic feature of the webpage text based on the URL. ### Claim 12 (depends on 9) The system of claim 9, in which the target text comprises a user text which comprises a traffic feature identifying source of the user text. ### Claim 13 (depends on 12) The system of claim 12, in which the user text is associated with a user identity and comprises at least one of the following: a historical search query request item associated with the user identity, a microblog tag associated with the user identity, a webpage title browsed associated with the user identity. ### Claim 14 A non-transitory computer-readable storage medium storing instructions thereon for execution by at least one processing circuit for pushing content items, the instructions comprising: obtaining a target text and analyzing lexically the target text; extracting features, comprising traffic feature, from the target text; classifying hierarchically, the features extracted from the target text, based on a pre-trained hierarchical classification model considering traffic feature to obtain a hierarchical classification of the target text; selecting one or more among a plurality of content items based on the hierarchical classification of the target text; obtaining the selected content items; and pushing the obtained content items to a terminal; wherein the target text comprises a webpage text which comprises a traffic feature identifying source of the webpage text; wherein the plurality of content items comprise a plurality of promotion texts; wherein for each of the plurality of promotion texts: obtaining the promotion text and analyzing lexically the promotion text; extracting features, comprising traffic feature, from the promotion text; classifying hierarchically, the features extracted from the promotion text, based on a pre-trained hierarchical classification model considering traffic feature to obtain a hierarchical classification of the promotion text; and calculating similarity between the webpage text and the promotion text based on the hierarchical classification of the promotion text and the hierarchical classification of the webpage text, wherein the selecting one or more among the plurality of promotion texts is based on the similarity between the webpage text and each of the plurality of promotion texts.