Autonomous real time publishing

Abstract

Techniques for autonomous and automatic real-time publishing of content are described. In an example embodiment, one or more topic terms are obtained. A set of information that is related to the one or more topic terms is automatically acquired. Linguistic analysis on the set of information is automatically performed to determine a set of linguistic structures that are represented in the set of information. The set of linguistic structures is used to automatically create a set of content items that are responsive to searches that include the one or more topic terms. New content that includes the set of content items is then automatically published.

What is claimed is: 
     
         1 . A method comprising:
 a computing device determining one or more topic terms representing a search topic that is undersupplied with content in one or more online domains;   the computing device acquiring a set of information that is related to the one or more topic terms;   the computing device performing linguistic analysis on the set of information to determine a set of linguistic structures that are represented in the set of information, wherein each linguistic structure in the set of linguistic structures includes a data element and one or more identifiers that respectively identify one or more linguistic categories which are associated with the data element;   the computing device determining a set of relevant linguistic structures that are represented in the set of information;   the computing device using the set of relevant linguistic structures to create a set of content items that are responsive to searches that include the one or more topic terms based on the data element included in each linguistic structure of the set of relevant linguistic structures;   the computing device creating new content based on the set of content items;   the computing device publishing the new content to the one or more online domains.   
     
     
         2 . The method of  claim 1 , wherein determining the set of relevant linguistic structures that are represented in the set of information is performed by at least:
 generating one or more semantic queries based on a semantic query template and the one or more topic terms,   executing the one or more semantic queries on the set of linguistic structures based on, for each linguistic structure in the set of linguistic structures, how relevant the linguistic structure is to the one or more topic terms to generate the set of relevant linguistic structures.   
     
     
         3 . The method of  claim 1 , wherein determining the one or more topic terms is based on one or more of:
 published search trends of one or more search engines, one or more internet service providers, or one or more web tracking services,   internal web traffic logs or search query logs,   trends or frequencies of new keywords or concepts appearing on news sites, blogs, or social media streams, or   cyclical trends or frequencies of keywords or concepts derived from data mining time stamped data sources.   
     
     
         4 . The method of  claim 1 , wherein determining the one or more topic terms is performed by at least:
 receiving search volume information from one or more data sources related to one or more search queries, wherein the search volume information is time stamped with particular time periods for each of the one or more search queries;   receiving search engine optimization competition information that reflects how much content is available via search engines for the one or more search queries;   generating time series data based on the search volume information and the search engine optimization competition information;   using the time series data to generate a set of candidate topic terms, a probability distribution of search volumes over time intervals for each topic associated with each candidate topic term of the set of the candidate topic terms, and an associated expectation value for search volumes for each candidate topic term of the set of candidate topic terms in a given time interval;   estimating a number of expected accesses to content related to one or more candidate search queries that each include one or more candidate topic terms of the set of candidate topic terms for one or more time intervals based on the probability distribution and associated expectation value of candidate topic terms included in each of the one or more candidate search queries;   generating the one or more topic terms based on the number of expected accesses to content related to the one or more candidate search queries for the one or more time intervals.   
     
     
         5 . The method of  claim 4 , wherein the one or more data sources include at least two disparate data sources. 
     
     
         6 . The method of  claim 1 , wherein the one or more online domains relate to content including one or more of: one or more web sites, one or more question and answering services, or social media. 
     
     
         7 . The method of  claim 1 , wherein the set of information comprises text items, and wherein performing linguistic analysis comprises:
 for each text item in the set of information:   parsing the text item into a set of words;   determining one or more part-of-speech linguistic structures for the set of words;   determining one or more phrasal linguistic structures based at least on the one or more part-of-speech linguistic structures;   determining one or more semantic-role linguistic structures based at least on the one or more phrasal linguistic structures;   storing, as part of the set of linguistic structures, one or more of the one or more part-of-speech linguistic structures, the one or more phrasal linguistic structures, and the one or more semantic-role linguistic structures.   
     
     
         8 . The method of  claim 7 , wherein performing linguistic analysis further comprises:
 determining one or more entity linguistic structures based on the one or more part-of-speech linguistic structures, wherein the one or more entity linguistic structures include one or more identifiers of entity categories;   normalizing the one or more entity linguistic structures by determining and assigning, to at least one of the one or more entity linguistic structures, one or more normalization values;   storing, as part of the set of linguistic structures, one or more of the one or more entity linguistic structures and the one or more normalization values.   
     
     
         9 . The method of  claim 7 , wherein:
 the one or more part-of-speech linguistic structures include one or more identifiers of linguistic categories that include one or more of: a proper name category, a verb group category, a determiner category, a noun category, a prepositional category, and a data context category;   the one or more phrasal linguistic structures include one or more identifiers of linguistic categories that include one or more of: a noun phrase category, a verb phrase category, and a prepositional phrase category;   the one or more semantic-role linguistic structures include one or more identifiers of semantic roles that include one or more of: a subject role, a predicate role, an object role, a temporal role, and a location role.   
     
     
         10 . The method of  claim 1 , further comprising: based on the set of relevant linguistic structures, automatically creating one or more natural language titles for the new content. 
     
     
         11 . The method of  claim 1 , further comprising:
 creating one or more natural language elements for the new content based on the set of relevant linguistic structures, wherein the one or more natural language elements include one or more of: one or more text statements, one or more media objects, or a conversational dialog.   
     
     
         12 . The method of  claim 11 , wherein the one or more natural language elements include the one or more text statements and the one or more text statements are created by natural language generation configured to generate the one or more text statements based on the one or more identifiers associated with the set of relevant linguistic structures. 
     
     
         13 . The method of  claim 11 , wherein the one or more natural language elements includes the one or more media objects and the one or more media objects are created using a discourse planner that generates output representing a dialog between characters and an animation rendering agent that uses text-to-speech to automatically render the dialog as a full motion video. 
     
     
         14 . The method of  claim 11 , wherein the one or more natural language elements includes the conversational dialog and the conversational dialog is created by combining statements or sentences emitted by a chatterbot program together with statements created from a natural language generator. 
     
     
         15 . The method of  claim 11 , wherein:
 the one or more text statements include one or more of: one or more comments, one or more opinions, one or more questions, or one or more answers,   the media objects include one or more of: one or more images, one or more audio clips, or one or more video object,   the conversational dialog includes one or more of: dialog between a questioner and an answerer, a blogger providing comments, or opposite sides of a debate.   
     
     
         16 . A non-transitory computer-readable storage medium storing one or more instructions which, when executed by one or more processors, cause the one or more processors to perform steps comprising:
 a computing device determining one or more topic terms representing a search topic that is undersupplied with content in one or more online domains;   the computing device acquiring a set of information that is related to the one or more topic terms;   the computing device performing linguistic analysis on the set of information to determine a set of linguistic structures that are represented in the set of information, wherein each linguistic structure in the set of linguistic structures includes a data element and one or more identifiers that respectively identify one or more linguistic categories which are associated with the data element;   the computing device determining a set of relevant linguistic structures that are represented in the set of information;   the computing device using the set of relevant linguistic structures to create a set of content items that are responsive to searches that include the one or more topic terms based on the data element included in each linguistic structure of the set of relevant linguistic structures;   the computing device creating new content based on the set of content items;   the computing device publishing the new content to the one or more online domains.   
     
     
         17 . The non-transitory computer-readable storage medium of  claim 16 , wherein determining the set of relevant linguistic structures that are represented in the set of information is performed by at least:
 generating one or more semantic queries based on a semantic query template and the one or more topic terms,   executing the one or more semantic queries on the set of linguistic structures based on, for each linguistic structure in the set of linguistic structures, how relevant the linguistic structure is to the one or more topic terms to generate the set of relevant linguistic structures.   
     
     
         18 . The non-transitory computer-readable storage medium of  claim 16 , wherein determining the one or more topic terms is based on one or more of:
 published search trends of one or more search engines, one or more internet service providers, or one or more web tracking services,   internal web traffic logs or search query logs,   trends or frequencies of new keywords or concepts appearing on news sites, blogs, or social media streams, or   cyclical trends or frequencies of keywords or concepts derived from data mining time stamped data sources.   
     
     
         19 . The non-transitory computer-readable storage medium of  claim 16 , wherein determining the one or more topic terms is performed by at least:
 receiving search volume information from one or more data sources related to one or more search queries, wherein the search volume information is time stamped with particular time periods for each of the one or more search queries;   receiving search engine optimization competition information that reflects how much content is available via search engines for the one or more search queries;   generating time series data based on the search volume information and the search engine optimization competition information;   using the time series data to generate a set of candidate topic terms, a probability distribution of search volumes over time intervals for each topic associated with each candidate topic term of the set of the candidate topic terms, and an associated expectation value for search volumes for each candidate topic term of the set of candidate topic terms in a given time interval;   estimating a number of expected accesses to content related to one or more candidate search queries that each include one or more candidate topic terms of the set of candidate topic terms for one or more time intervals based on the probability distribution and associated expectation value of candidate topic terms included in each of the one or more candidate search queries;   generating the one or more topic terms based on the number of expected accesses to content related to the one or more candidate search queries for the one or more time intervals.   
     
     
         20 . The non-transitory computer-readable storage medium of  claim 19 , wherein the one or more data sources include at least two disparate data sources. 
     
     
         21 . The non-transitory computer-readable storage medium of  claim 16 , wherein the one or more online domains relate to content including one or more of: one or more web sites, one or more question and answering services, or social media. 
     
     
         22 . The non-transitory computer-readable storage medium of  claim 16 , wherein the set of information comprises text items, and wherein performing linguistic analysis comprises:
 for each text item in the set of information:   parsing the text item into a set of words;   determining one or more part-of-speech linguistic structures for the set of words;   determining one or more phrasal linguistic structures based at least on the one or more part-of-speech linguistic structures;   determining one or more semantic-role linguistic structures based at least on the one or more phrasal linguistic structures;   storing, as part of the set of linguistic structures, one or more of the one or more part-of-speech linguistic structures, the one or more phrasal linguistic structures, and the one or more semantic-role linguistic structures.   
     
     
         23 . The non-transitory computer-readable storage medium of  claim 22 , wherein performing linguistic analysis further comprises:
 determining one or more entity linguistic structures based on the one or more part-of-speech linguistic structures, wherein the one or more entity linguistic structures include one or more identifiers of entity categories;   normalizing the one or more entity linguistic structures by determining and assigning, to at least one of the one or more entity linguistic structures, one or more normalization values;   storing, as part of the set of linguistic structures, one or more of the one or more entity linguistic structures and the one or more normalization values.   
     
     
         24 . The non-transitory computer-readable storage medium of  claim 22 , wherein:
 the one or more part-of-speech linguistic structures include one or more identifiers of linguistic categories that include one or more of: a proper name category, a verb group category, a determiner category, a noun category, a prepositional category, and a data context category;   the one or more phrasal linguistic structures include one or more identifiers of linguistic categories that include one or more of: a noun phrase category, a verb phrase category, and a prepositional phrase category;   the one or more semantic-role linguistic structures include one or more identifiers of semantic roles that include one or more of: a subject role, a predicate role, an object role, a temporal role, and a location role.   
     
     
         25 . The non-transitory computer-readable storage medium of  claim 16 , further comprising: based on the set of relevant linguistic structures, automatically creating one or more natural language titles for the new content. 
     
     
         26 . The non-transitory computer-readable storage medium of  claim 16 , further comprising:
 creating one or more natural language elements for the new content based on the set of relevant linguistic structures, wherein the one or more natural language elements include one or more of: one or more text statements, one or more media objects, or a conversational dialog.   
     
     
         27 . The non-transitory computer-readable storage medium of  claim 26 , wherein the one or more natural language elements include the one or more text statements and the one or more text statements are created by natural language generation configured to generate the one or more text statements based on the one or more identifiers associated with the set of relevant linguistic structures. 
     
     
         28 . The non-transitory computer-readable storage medium of  claim 26 , wherein the one or more natural language elements includes the one or more media objects and the one or more media objects are created using a discourse planner that generates output representing a dialog between characters and an animation rendering agent that uses text-to-speech to automatically render the dialog as a full motion video. 
     
     
         29 . The non-transitory computer-readable storage medium of  claim 26 , wherein the one or more natural language elements includes the conversational dialog and the conversational dialog is created by combining statements or sentences emitted by a chatterbot program together with statements created from a natural language generator. 
     
     
         30 . The non-transitory computer-readable storage medium of  claim 26 , wherein:
 the one or more text statements include one or more of: one or more comments, one or more opinions, one or more questions, or one or more answers,   the media objects include one or more of: one or more images, one or more audio clips, or one or more video object,   the conversational dialog includes one or more of: dialog between a questioner and an answerer, a blogger providing comments, or opposite sides of a debate.