Generating personalized content

Abstract

Systems and methods are presented for adding system-identified content to a user's content space based on information identified from the user's email account. In operation, after obtaining user authorization to access email items of an email account, the email items are iteratively processed. In processing each of the email items, one or more topics of a currently processed email item are identified. Based on the identified topics and user preferences, one or more content items may be identified for addition to the user's online content space. Additionally, the information identified from processing the one or more email items is used to update the user's preferences.

What is claimed: 
     
         1 . A computing system, comprising:
 one or more processors; and   a memory storing program instructions that, when executed by the one or more processors, cause the one or more processors to at least:
 obtain user authorization for accessing a plurality of items associated with a user of a social networking service; 
 access, via a plurality of calls to an application programming interface (API), the plurality of items, wherein the API provides information associated with the plurality of items over a computer network to the social networking service in response to the plurality of calls to the API without providing the plurality of items to the social networking service; 
 process the information associated with the plurality of items to:
 determine a topic of a first item of the plurality of items; 
 determine, using a user preference manager and based at least in part on the topic, a new user preference or a user preference update, wherein the user preference manager includes a trained machine learning model trained to determine at least one of user preferences or updates to user preferences based at least in part on input data that includes user information; and 
 determine, based at least in part on the topic and the new user preference or the user preference update, a content item in connection with the user. 
 
   
     
     
         2 . The computing system of  claim 1 , wherein the program instructions, when executed by the one or more processors, further cause the one or more processors to:
 determine a strength value associated with the topic;   determine a sentiment value associated with the topic;   determine, based at least in part on the strength value and the sentiment value, a topic score;   determine that the topic score exceeds a first threshold value; and   determine, in response to the determination that the topic score exceeds the first threshold value, that the content item is to be added to a content space associated with the user.   
     
     
         3 . The computing system of  claim 2 , wherein the program instructions, when executed by the one or more processors, further cause the one or more processors to:
 determine a similarity between the content item and a first content collection of the content space associated with the user, wherein the similarity is determined based on at least one of:
 a first comparison of the topic with a collection topic associated with the first content collection; or 
 a second comparison of the content item with a plurality of content items associated with the first content collection; 
   determine that the similarity exceeds a second threshold value; and   in response to the determination that the similarity exceeds the second threshold value, add the content item to the first content collection of the content space associated with the user.   
     
     
         4 . The computing system of  claim 2 , wherein the program instructions, when executed by the one or more processors, further cause the one or more processors to:
 determine a similarity between the content item and a first content collection of the content space associated with the user, wherein the similarity is determined based on at least one of:
 a first comparison of the topic with a collection topic associated with the first content collection; or 
 a second comparison of the content item with a plurality of content items associated with the first content collection; 
   determine that the similarity does not exceed a second threshold value; and   in response to the determination that the similarity does not exceed the second threshold value:
 create a second content collection in the content space associated with the user; and 
 add the content item to the second content collection of the content space associated with the user. 
   
     
     
         5 . The computing system of  claim 1 , wherein the plurality of items includes at least one of:
 an email message;   a task item;   a calendar item;   a note item;   a short messenger service message (SMS) item; or   a multimedia message service (MMS) message item.   
     
     
         6 . A computing-implemented method, comprising:
 obtaining user authorization for accessing a plurality of messages associated with a user of a social networking service;   accessing, via a call to an application programming interface (API), a first message of the plurality of messages, wherein the API provides information associated with the first message over a computer network to the social networking service in response to the call to the API without providing the first message to the social networking service;   determining a topic of the first message;   determining, using a user preference engine and based at least in part on the topic, a new user preference of the user or a user preference update of the user; and   determining, based at least in part on the topic and the new user preference or the user preference update, a content item in connection with the user.   
     
     
         7 . The computer implemented method of  claim 6 , wherein determining the topic of the first message includes at least one of:
 performing a lexical analysis of the first message to identify key terms of the first message to determine the topic of the first message; or   processing, using a trained machine learning model trained to identify one or more message topics based at least in part on input data that includes a message, the first message to determine the topic of the first message.   
     
     
         8 . The computer-implemented method of  claim 6 , wherein determining the topic of the first message includes:
 determining at least one of a template or a data structure associated with data included in the first message; and   performing an analysis of the data included in the first message based at least in part on the template or the data structure to determine the topic of the first message.   
     
     
         9 . The computer-implemented method of  claim 6 , wherein determining the topic of the first message includes:
 identifying, in the first message, a reference to additional content;   processing the additional content to determine the topic of the first message.   
     
     
         10 . The computer-implemented method of  claim 9 , wherein the additional content includes at least one of:
 a uniform resource indicator;   a uniform resource locator;   a message thread associated with the first message; or   a second message of the plurality of messages.   
     
     
         11 . The computer-implemented method of  claim 6 , further comprising:
 determining a strength value associated with the topic;   determining a sentiment value associated with the topic;   determining, based at least in part on the strength value and the sentiment value, a topic score;   determining that the topic score exceeds a first threshold value; and   determining, in response to the determination that the topic score exceeds the first threshold value, that the content item is to be added to a content space associated with the user.   
     
     
         12 . The computer-implemented method of  claim 11 , further comprising:
 determining, a similarity between the content item and a first content collection of the content space associated with the user, wherein the similarity is determined based on at least one of:
 a first comparison of the topic with a collection topic associated with the first content collection; or 
 a second comparison of the content item with a plurality of content items associated with the first content collection; 
   determining that the similarity exceeds a second threshold value; and   in response to the determination that the similarity exceeds the second threshold value, adding the content item to the first content collection of the content space associated with the user.   
     
     
         13 . The computer-implemented method of  claim 11 , further comprising:
 determining, a similarity between the content item and a first content collection of the content space associated with the user, wherein the similarity is determined based on at least one of:
 a first comparison of the topic with a collection topic associated with the first content collection; or 
 a second comparison of the content item with a plurality of content items associated with the first content collection; 
   determining that the similarity does not exceed a second threshold value; and   in response to the determination that the similarity does not exceed the second threshold value:
 creating a second content collection to the content space associated with the user; and 
 adding the content item to the second content collection of the content space associated with the user. 
   
     
     
         14 . The computer-implemented method of  claim 6 , wherein:
 determining the topic of the first message includes identifying an item purchased by the user; and   determining the content item includes determining that the content item is complementary to the item purchased by the user.   
     
     
         15 . The computer-implemented method of  claim 14 , wherein:
 determining the topic of the first message includes identifying a plurality of items purchased by the user;   determining the content item includes determining a plurality of content items that are complementary to the plurality of items purchased by the user; and   the method further comprises:
 creating a new content collection including the plurality of items purchased by the user and the plurality of content items. 
   
     
     
         16 . A computing system, comprising:
 one or more processors; and   a memory storing program instructions that, when executed by the one or more processors, cause the one or more processors to at least:
 obtain user authorization for accessing a plurality of messages associated with a user of a social networking service; 
 access, via a plurality of calls to an application programming interface (API), the plurality of messages, wherein the API provides information associated with the plurality of messages over a computer network to the social networking service in response to the plurality of calls to the API without providing the plurality of messages to the social networking service; 
 process the information associated with the plurality of messages to:
 determine a topic of a first message of the plurality of messages; and 
 determine, using a user preference manager and based at least in part on the topic, a new user preference or a user preference update, wherein the user preference manager includes a trained machine learning model trained to determine at least one of user preferences or updates to user preferences based at least in part on input data that includes user information. 
 
   
     
     
         17 . The computing system of  claim 16 , wherein the program instructions, when executed by the one or more processors, further cause the one or more processors to:
 determine, based at least in part on at least one of the topic, the new user preference, or the user preference update, a content item to be added to a content space associated with the user.   
     
     
         18 . The computing system of  claim 17 , wherein the program instructions, when executed by the one or more processors, further cause the one or more processors to:
 based at least in part on a comparison of a plurality of content collections of the content space associated with the user to at least one of the topic or the content item, at least one of:
 add the content item to a first content collection of the plurality of content collections; or 
 create a new content collection in the content space associated with the user and add the content item to a new content collection. 
   
     
     
         19 . The computing system of  claim 16 , wherein determination of the topic of the first message includes:
 determining at least one of a template or a data structure associated with data included in the first message; and   performing an analysis of the data included in the first message based at least in part on the template or the data structure to determine the topic of the first message.   
     
     
         20 . The computing system of  claim 16 , wherein determination of the topic of the first message includes:
 determining a strength value associated with the topic; and   determining a sentiment value associated with the topic.