Abstract: Techniques for identifying and presenting member profiles similar to a source member profile are described. With some embodiments, a general recommendation engine is used to extract features from member profiles, and then store the extracted features, including any computed, derived or retrieved profile features, in an enhanced member profile. In real-time, the general recommendation engine processes client requests to identify member profiles similar to a source member profile by comparing select profile features stored in the enhanced member profile with corresponding profile features of the source member profile, where the comparison results in several similarity sub-scores that are then combined in accordance with directives set forth in a configuration file. Finally, the member profiles with the highest similarity scores corresponding with the user-selected member profile are selected, and in some instances, presented to a user.

Abstract: A machine may be configured to detect an anomalous event based on metrics pertaining to a production system. For example, the machine analyzes a time series of values associated with a metric pertaining to a production system. The machine identifies a pattern associated with the time series based on the analysis of the time series. The pattern may describe an occurrence of particular values at particular timestamps of the time series. The machine determines a range of potential values for a next timestamp in the time series based on the pattern. The machine assigns a score value to an actual value associated with the metric and corresponding to the next timestamp. The assigning of the score value may be based on a comparison of the actual value and the range of potential values. The machine identifies the actual value as a candidate for an alert based on the score value.

Abstract: Techniques for identifying and presenting member profiles similar to a source member profile are described. With some embodiments, a general recommendation engine is used to extract features from member profiles, and then store the extracted features, including any computed, derived or retrieved profile features, in an enhanced member profile. In real-time, the general recommendation engine processes client requests to identify member profiles similar to a source member profile by comparing select profile features stored in the enhanced member profile with corresponding profile features of the source member profile, where the comparison results in several similarity sub-scores that are then combined in accordance with directives set forth in a configuration file. Finally, the member profiles with the highest similarity scores corresponding with the user-selected member profile are selected, and in some instances, presented to a user.

Abstract: Systems and methods are disclosed that recommend one or more electronic presentations to a user based on one or more factors. These factors may include contextual information, behavioral information, profile information, or combinations of the foregoing. Contextual information may include content and/or features extracted from a given electronic presentation. Behavioral information may include user behavioral data, such as the number of times a user has viewed a presentation, the amount of the presentation viewed by the user, presentations previously viewed by the user, and other such behavioral data. Profile information may include user professional profile information, such as skills the user has identified as possessing, employment history information, and other such user professional profile information.

Abstract: A first user of a presentation machine may be a recruiter that initiates an action in reference to a first user profile. The first user profile may describe a first candidate for a job. The presentation machine may identify the first user profile and determine a similarity score that indicates a degree of similarity between the first user profile and a second user profile that describes a second candidate for the job. The presentation machine may also access a volatility score that indicates a likelihood that the second candidate is receptive to a proposal that the second candidate be employed by an employer. Based on the similarity score and on the volatility score, the presentation machine may determine a rank of the second user profile. Based on the determined rank, the presentation machine may present the second user profile to the first user.

Abstract: Techniques for identifying and presenting member profiles similar to a source member profile are described. With some embodiments, a general recommendation engine is used to extract features from member profiles, and then store the extracted features, including any computed, derived or retrieved profile features, in an enhanced member profile. In real-time, the general recommendation engine processes client requests to identify member profiles similar to a source member profile by comparing select profile features stored in the enhanced member profile with corresponding profile features of the source member profile, where the comparison results in several similarity sub-scores that are then combined in accordance with directives set forth in a configuration file. Finally, the member profiles with the highest similarity scores corresponding with the user-selected member profile are selected, and in some instances, presented to a user.

Abstract: The disclosed embodiments provide a system and method for performing version control for asynchronous distributed machine learning. During operation, the system transmits a first global version of a statistical model to a set of client computer systems. Next, the system obtains, from a first subset of the client computer systems, a first set of updates to the first global version. The system then merges the first set of updates into a second global version of the statistical model. Finally, the system transmits the second global version to the client computer systems asynchronously from receiving a second set of updates to the first and/or second global versions from a second subset of the client computer systems.

Abstract: The disclosed embodiments provide a method and system for performing regularized model adaptation for in-session recommendations. During operation, the system obtains, from a server, a first global version of a statistical model. During a first user session with a user, the system improves a performance of the statistical model by using the first global version to output one or more recommendations to the user and using the first global version and user feedback from the user to create a first personalized version of the statistical model. At an end of the first user session, the system transmits an update containing a difference between the first personalized version and the first global version to the server for use in producing a second global version of the statistical model by the server.

Abstract: A first user of a presentation machine may be a recruiter that initiates an action in reference to a first user profile. The first user profile may describe a first candidate for a job. The presentation machine may identify the first user profile and determine a similarity score that indicates a degree of similarity between the first user profile and a second user profile that describes a second candidate for the job. The presentation machine may also access a volatility score that indicates a likelihood that the second candidate is receptive to a proposal that the second candidate be employed by an employer. Based on the similarity score and on the volatility score, the presentation machine may determine a rank of the second user profile. Based on the determined rank, the presentation machine may present the second user profile to the first user.

Abstract: Disclosed in some examples are methods, systems and machine readable medium for recommending an out-of-network communication by determining a set of potential recommended members of a social networking service based upon one or more recommendation criteria. In some examples the recommendation criteria may include: a profile similarity to a previous target of an out-of-network communication, a degree of correspondence between an interest and intent of the sending member, and a likelihood of response.

Abstract: Techniques for identifying and presenting member profiles similar to a source member profile are described. With some embodiments, a general recommendation engine is used to extract features from member profiles, and then store the extracted features, including any computed, derived or retrieved profile features, in an enhanced member profile. In real-time, the general recommendation engine processes client requests to identify member profiles similar to a source member profile by comparing select profile features stored in the enhanced member profile with corresponding profile features of the source member profile, where the comparison results in several similarity sub-scores that are then combined in accordance with directives set forth in a configuration file. Finally, the member profiles with the highest similarity scores corresponding with the user-selected member profile are selected, and in some instances, presented to a user.

Abstract: The disclosed embodiments relate to a system that uses data from an online social network to optimize subscription offers. During operation of the online social network, the system gathers data associated with subscription offers that were presented to members of the online social network, including information about which subscription offers were converted. Next, the system uses a machine-learning technique to train a model based on the gathered data. Finally, the system uses the trained model to select subscription offers to present to a member of the online social network.

Abstract: A statistically overrepresented token in the descriptions of users associated with a target entity may be descriptive of the target entity. This may be true regardless of whether a primary description of the entity includes the overrepresented token. Accordingly, the entity description machine may access multiple descriptions of multiple users associated with the target entity. A portion of the multiple descriptions may each include a token descriptive of the target entity and of a subset of the multiple users. The entity description machine may determine that the token is overrepresented among the tokens within the multiple descriptions and generate a supplemental description of the target entity, where the supplemental description includes the overrepresented token. Once the supplemental description is generated, the entity description machine may use the supplemental description in referencing the target entity.

Abstract: Disclosed in some examples are methods, systems and machine readable medium for recommending an out-of-network communication by determining a set of potential recommended members of a social networking service based upon one or more recommendation criteria. In some examples the recommendation criteria may include: a profile similarity to a previous target of an out-of-network communication, a degree of correspondence between an interest and intent of the sending member, and a likelihood of response.

Abstract: Techniques for assisting a user in determining an affinity between a candidate and an organization. According to various embodiments, company data is received and includes a set of position characteristics and a set of pool characteristics. Member data is received and includes a set of member characteristics. A set of member characteristic scores are generated. Each characteristic score is based on comparing a member characteristic of the set of member characteristics with a position characteristic of the set of position characteristics and a pool characteristic of the set of pool characteristics. A member fit score is determined based on the set of member characteristic scores. A relative fit score is generated for the member based on a comparison of the member fit score and a set of second member fit scores for a second set of members. An identification of an organization is presented based on the relative fit score.

Abstract: Techniques for assisting a user in determining an interest level between a member of a social network system and an organization. According to various embodiments, applicant data is accessed for applicants having applied to an organization. A set of common applicant characteristics is determined for the set of applicant data. Member data is accessed indicative of a member of an online social media network. An interest score is generated based on a comparison of the member data and the set of applicant data. An identification of the organization is presented based on the interest score.

Abstract: The disclosed embodiments provide a system for performing network A/B testing. During operation, the system obtains a graph of a social network and calculates a set of equally sized clusters of users in the social network by iteratively switching memberships of the nodes among the equally sized clusters to increase a number of edges in each of the equally sized clusters. Next, the system randomly selects a subset of the equally sized clusters for exposure to a treatment version of a message. The system then performs an A/B test by presenting the treatment version to the selected clusters and tracking a response of the selected clusters to the treatment version.

Abstract: The disclosed embodiments provide a method and system for performing network A/B testing. During operation, the system obtains, for a set of users in a social network, a set of treatment assignments of the users in an A/B test, wherein the treatment assignments indicate exposure of the users to a control version or a treatment version of a message. Next, the system obtains, for each of the users, a fraction of neighbors exposed to the treatment version in the A/B test. The system then applies a statistical model to the treatment assignments and the fraction of neighbors exposed to the treatment version to estimate an average treatment effect (ATE) for the set of users. Finally, the system selects, based on the ATE, a fraction of additional users in the social network for subsequent exposure to the treatment version and presents the treatment version to the fraction of additional users.

Abstract: A statistically overrepresented token in the descriptions of users associated with a target entity may be descriptive of the target entity. This may be true regardless of whether a primary description of the entity includes the overrepresented token. Accordingly, the entity description machine may access multiple descriptions of multiple users associated with the target entity. A portion of the multiple descriptions may each include a token descriptive of the target entity and of a subset of the multiple users. The entity description machine may determine that the token is overrepresented among the tokens within the multiple descriptions and generate a supplemental description of the target entity, where the supplemental description includes the overrepresented token. Once the supplemental description is generated, the entity description machine may use the supplemental description in referencing the target entity.

Abstract: A machine may implement a recommender that provides recommendations to users. The machine may be configured to present a first version of the recommender configured by various parameters. A user may submit a message to the machine, and the machine may identify a parameter among the various parameters by tokenizing the message and identifying the parameter among the tokens. The machine may then generate a second version of the recommender by modifying the parameter and configuring the second version according to the modified parameter. The machine may then present the first and second versions of the recommender contemporaneously two different portions of the users. By tokenizing a further message received from the users, the machine may evaluate the first and second versions and determine whether the second version is a replacement of the first version.