# US 2018/0322206 A1 · Personalized user-categorized recommendations > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 2018/0322206 A1 - Document ID: US20180322206A1 - Status: Published application - Kind: A1 - Application number: 15/627,963 - Filed: 2017-06-20 - Published: 2018-11-08 - Earliest claimed: 2017-05-05 - Assignees: Microsoft Technology Licensing, LLC - Applicants: Microsoft Technology Licensing, LLC - Inventors: Yanfeng Sun; Alexandre Bernard Raymond Michelis; Arun Kumar Sacheti; Emily Lin; Ryuichi Hirano; Vincent Kin-Wah Leung - Web page: https://patentgrep.com/patent/US20180322206A1 ## CPC - G06F 16/9535: Search customisation based on user profiles and personalisation - G06F 16/9024: {Graphs; Linked lists (G06F16/9027 takes precedence)} - G06F 16/9038: Presentation of query results - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) ## Abstract Methods and systems for providing targeted recommendations are provided. A user model is generated from user data. Feed candidates are generated based on the user model. The generated feed candidates are ranked on a predetermined scale. At least one targeted recommendation from the ranked feed candidates. The at least one targeted recommendation feed is provided to a user associated with the user data. ## Claims ### Claim 1 A computer-implemented method for providing targeted recommendations, comprising: generating a user model from user data associated with a user; determining feed candidates based on the user model, wherein determining the feed candidates includes determining a topic specific model from the user model and performing a search based on the topic specific model; determining at least one targeted recommendation from the feed candidates, wherein determining the at least one targeted recommendation includes filtering search results based on the user model; and providing the at least one targeted recommendation to the user. ### Claim 2 (depends on 1) The method of claim 1, wherein generating the user model comprises generating one or more user vectors from the user data. ### Claim 3 (depends on 2) The method of claim 2, wherein one or more user vectors comprises: mapping the user data to text; generating one or more bigrams from the text; and generating the one or more user vectors from the bigrams. ### Claim 4 (depends on 2) The method of claim 2, wherein generating the user model comprises generating a self-organizing map comprising the one or more user vectors. ### Claim 5 (depends on 4) The method of claim 4, further comprising dynamically updating the self-organizing map. ### Claim 6 (depends on 1) The method of claim 1, wherein updating the self-organizing map comprises including at least one new user vector into the self-organizing map, the at least one new user vector corresponding to a new interest associated with the user. ### Claim 7 (depends on 1) The method of claim 1, further comprising ranking the feed candidates on a predetermined scale. ### Claim 8 (depends on 1) The method of claim 1, wherein the user data comprises a user category created by a user. ### Claim 9 (depends on 8) The method of claim 8, further comprising discontinuing providing the at least one targeted recommendation upon deletion of the user category. ### Claim 10 A system for providing targeted recommendations, the system comprising: at least one processing device; and at least one computer readable data storage device storing instructions that, when executed by the at least one processing device, cause the at least one processing device to: receive user data comprising one or more user interests from a user device; generate one or more user models based on the one or more user interests; generate one or more topic specific models from one or more user models; perform a search based on the one or more topic specific models; filter search results based on the one or more user models to generate feed candidates; determine at least one targeted recommendation feed from the feed candidates; and provide the at least one targeted recommendation feed. ### Claim 11 (depends on 10) The system of claim 10, wherein instructions that, when the instructions executed by the at least one processing device, cause the at least one processing device to: rank the feed candidates on a predetermined scale; and determine the at least one targeted recommendation feed from the feed candidates based on the rank. ### Claim 12 (depends on 10) The system of claim 10, wherein the instructions when executed by the at least one processing device cause the at least one processing device to: generate a task specific model; generate a non-personalized topic feed based on the task-specific user model; and generate the feed candidates from the non-personalized topic feed based on the user model. ### Claim 13 (depends on 12) The system of claim 12, wherein the instructions when executed by the at least one processing device cause the at least one processing device to: map a user task from the user data to the task specific model. ### Claim 14 (depends on 10) The system of claim 10, wherein the instructions when executed by the at least one processing device cause the at least one processing device to: map the user data to text; generate one or more bigrams for the text; and generate the user model from the bigrams, the user model comprising one or more vectors generated from the one or more bigrams. ### Claim 15 A computer readable storage device including computer readable instructions, which when executed by a processing unit is operative to: receive user data comprising a user interest from a user device; generate a user model based on the user data, the user model comprising at least one user vector corresponding to the user interest; generate a topic specific model from the user model; perform a search based on the topic specific model; filter search results based on the user model to generate feed candidates; rank the generated feed candidates on a predetermined scale; determine at least one targeted recommendation from the feed candidates; and provide the at least one targeted recommendation feed to a user associated with the user device. ### Claim 16 (depends on 15) The computer readable storage device of claim 15, wherein the computer readable instructions when executed by the processing unit is further operative to: generate a user group comprising a plurality of users associated with a topic; generate a topic specific model; generate a non-personalized topic feed; and generate the feed candidates for the user from the non-personalized topic feeds. ### Claim 17 (depends on 16) The computer readable storage device of claim 16, wherein the computer readable instructions when executed by the processing unit is further operative to: map the user into the user group. ### Claim 18 (depends on 16) The computer readable storage device of claim 16, wherein the computer readable instructions when executed by the processing unit is further operative to: generate the feed candidates for the user from the non-personalized topic feeds by filtering the feed candidates from the non-personalized topic based on the at least user vectors. ### Claim 19 (depends on 15) The computer readable storage device of claim 15, wherein the computer readable instructions when executed by the processing unit is further operative to: create a self-organizing map comprising the at least one user vector. ### Claim 20 (depends on 19) The computer readable storage device of claim 19, wherein the computer readable instructions when executed by the processing unit is further operative to: dynamically update the self-organizing map based on a feedback from the user for the provided at least one targeted recommendation.