Facebook, Inc.

Facbook, Inc.

Rajat Raina

Sriram Sankar

Kihyuk Hong

Kittipat Virochsiri

Gintaras Andrius Woss

Kedar Dhamdhere

Maxim Gubin

Michael Curtiss

Chaitanya Mishra

Olivier Chatot

Zhongxian Chen

Xiao Li

Maxime Boucher

Allan Stewart

Li-Tal Mashiach

Kurchi Subhra Hazra

Ashish Gupta

Jinlong Zhou

Dinkar Jain

Jiajin Yu

Eric Ringger

Rose Marie Philip

Hetunandan Munisharan Kamichetty

In one embodiment, a method includes receiving a search query and identifying one or more n-grams. The method also includes accessing, for each identified n-gram, a time-based distribution that includes a count of actions associated with the n-gram. The actions may have occurred within several time windows, and the time-based distribution may count the actions within each time window. The method may also include determining one or more extrema time windows. The method may also include modifying the search query based on the determined extrema time windows. The method also includes executing the modified search query, ranking identified content objects, and sending search results to a client system. The search results may be displayed in ranked order based on the rankings of the corresponding content objects.

Time-window counters for search results on online social networks

In one embodiment, a method includes receiving, from a client system of a first user, a text query to search for posts of the online social network, the text query comprising one or more n-grams, searching a personalized index of keyword phrases associated with the first user to identify one or more keyword phrases matching one or more of the n-grams of the text query, the index of keyword phrases being based on content extracted from posts by one or more second users that have been viewed by the first user, and providing to the client system in response to receiving the text query, instructions for displaying a user interface comprising one or more of suggested queries to search for posts of the online social network, wherein each suggested query comprises one or more n-grams from the text query and one or more of the identified keyword phrases.

Social static ranking for search

In one embodiment, a method includes receiving a query comprising one or more n-grams, generating a plurality of query commands based on a parsing of the query input, wherein each query command comprises an inner query constraint and an outer query constraint, scoring the query commands based at least in part on a number of objects matching the inner query constraint of the respective query command, selecting a first query command based at least in part on the respective scores of the query commands, and executing the first query command to identify a plurality of objects based on: (1) a first set of objects matching the first inner query constraint and at least in part matching the first outer query constraint, and (2) a second set of objects matching the first outer query constraint.

Search query interactions

In one embodiment, a method includes receiving a query comprising one or more n-grams, generating a plurality of query commands based on a parsing of the query input, wherein each query command comprises a plurality of query constraints, scoring the plurality of query commands based at least in part on a number of objects matching each of the query constraints of the respective query command, selecting a first query command based at least in part on the respective scores of the query commands, wherein the first query command comprises an inverse constraint corresponding to a first query constraint that has previously been flagged as identifying greater than a threshold number of objects and one or more second query constraints, and executing the first query command to identify a plurality of objects matching the inverse constraint and the one or more second query constraints.

Using inverse operators for queries

In one embodiment, a method includes receiving, from a client system of a first user, a query comprising one or more n-grams, determining one or more search intents of the query based at least on whether one or more of the n-grams in the query match terms corresponding to a search intent indexed in a pattern-detection model, generating one or more search results based on the query, each search result corresponding to an object of a plurality of objects, and scoring the search results based on one or more of the search intents.

Search intent for queries

In one embodiment, a method includes accessing a data set including a list of objects matching a query command and a score for each of the listed objects, where the query command is generated by parsing a query using a parsing algorithm, and where the score for each of the listed objects is calculated based on a scoring algorithm. The method also includes generating multiple subsets of the data set, each subset including one or more of the listed objects, and calculating, for each subset, a measure of score-quality associated with the scores of the objects in the subset and a measure of CPU-power associated with an amount of processing power required for retrieving the objects in the subset. The method also includes revising the parsing algorithm based on a comparison of the measures of score-quality and the measures of CPU-power associated with one or more of the subsets.

Rewriting search queries on online social networks

In one embodiment, a method includes retrieving multiple posts containing each of the query terms after receiving a search query for posts from a user of an online social network, identifying, from an index of known phrases, one or more known phrases in the text of the retrieved posts, filtering the retrieved posts to generate a filtered set of posts, wherein the filtering comprises removing each post not containing at least one of the identified known phrases, computing features for each of the identified known phrases, calculating a score for each post in the filtered set based on the features of the identified known phrases, ranking the posts based on the calculated scores, and sending instructions, to the client system, for generating a search-results interface comprising references to one or more of the posts presented in ranked order.

Measuring phrase association on online social networks

In one embodiment, a method includes receiving, from a client system of a first user, a request to post content to the online social network, identifying a plurality of second users, wherein each second user is associated with a privacy setting, and wherein a percentage of the identified second users have a privacy setting likely to make the content posted by the first user accessible to the second user, determining, for each identified second user, whether the content posted by the first user is accessible to the second user, and sending, to one or more client systems of one or more of the second users, respectively, instructions for presenting the content posted by the first user, wherein the content is sent to second users determined to be able to access the content and is not sent to second users determined to be not able to access the content.

Approximate privacy indexing for search queries on online social networks

In one embodiment, a method includes receiving a query inputted by the user; identifying a set of objects matching the query; calculating, for each identified object, a plurality of category-scores corresponding to a plurality of categories, respectively, wherein each category-score is calculated based on a plurality of sub-scores corresponding to a plurality of scoring axes; categorizing each identified object into a category of the plurality of categories based on the category-scores for the identified object; and sending, to the client system in response to the query, one or more search results corresponding to one or more of the categorized objects for display, each search result referencing the respective categorized object, wherein the one or more categorized objects of the search results comprises objects categorized into one or more selected categories.

Categorizing objects for queries on online social networks

In one embodiment, a method includes receiving an unstructured text query to search for posts of the online social network. The method includes parsing the text query to identify one or more n-grams. The method includes searching an index of keyword phrases associated with the first user to identify one or more keyword phrases matching one or more of the n-grams of the text query. The index of keyword phrases is based on posts by one or more second users of the online social network. The method includes calculating a keyword score for each of the identified keyword phrases. The method includes generating one or more suggested queries. Each suggested query includes one or more n-grams identified from the text query and one or more identified keyword phrases. The method includes sending one or more of the suggested queries to search for posts of the online social network.

In one embodiment, one or move computing devices receive, from a client device of a first user, a query corresponding to a first node type of a plurality of node types. The computer devices search one or more search indices to identify one or more nodes of a social graph matching the query. The search indices maintain an index of a plurality of nodes of the first node type, each node of the plurality of nodes having a static rank which is based at least in part on a pre-calculated value for each node. The computer devices send, to the client device of the first user for display, a search-results interface responsive to the received query. The search-results interface includes reference to one or more nodes each corresponding to the first node type and having a static rank greater than a threshold rank.

In one embodiment, a method includes retrieving, from one or more data stores, one or more objects requested by an application accessed by a first user of an online social network, wherein each object is associated with a privacy setting, and wherein a percentage of the retrieved objects have a privacy setting likely to make the object visible to the first user, determining, for each retrieved object, a visibility of the object with respect to the first user, and sending, to the application accessed by the first user, one or more of the retrieved objects, wherein each sent object has a visibility that is visible to the first user, and wherein objects having a visibility that is not visible to the first user are not sent.

In one embodiment, a method includes receiving, from a client system of a first user, a query comprising one or more n-grams, parsing the one or more n-grams of the query to generate a query command, wherein the query command comprises an inner query constraint and an outer query constraint, identifying from a plurality of objects based on the query command (1) a first set of objects matching the inner query constraint and at least in part matching the outer query constraint, and (2) a second set of objects matching the outer query constraint, and generating one or more search results based on the first and second sets of objects, wherein each search result corresponds to an object of the plurality of objects.

In one embodiment, a method includes receiving, from a client system of a first user, a query comprising one or more n-grams, parsing the one or more n-grams of the query to identify a first query constraint and one or more second query constraints, wherein the first query constraint has been previously flagged as identifying greater than a threshold number of objects, generating a query command based on the query, wherein the query command comprises an inverse constraint associated with the first query constraint and the one or more second query constraints, and generating one or more search results based on the query command, each search result corresponding to an object of a plurality of objects.

Growing community of inventors

Mountain View, CA, United States of America

Rajat Raina