Abstract: In an example embodiment, a machine-learned model is trained to specifically identify anomaly points in time series data. The model is capable of being applied in parallel to many different time series simultaneously, allowing for a scalable solution for large scale online networks. The model classifies each data point in a specified time window and outputs rich contextual information for downstream applications, such as ranking and display of the anomalous data points.

Abstract: In an example embodiment, a model is trained to specifically identify reversal points in data and then to rank these reversal points in order of importance. A reversal point shall be defined as a point in which a particular metric, specifically a first order derivative, crosses over from positive to negative or vice-versa. Users are more likely to be interested in abnormal and significant changes in data, and thus the machine-learned model is trained to evaluate a reversal point based on two dimensions: abnormality and significance.

Abstract: In an embodiment, the disclosed technologies include a method for generating a typeahead suggestion list for an input field of a search interface, including receiving, as digital input, a query string that has been extracted from the input field and context data that comprises one or more of: a search term that has been extracted from another input field of the search interface or a search criterion that has been extracted from a member profile that is associated with the query string via an online connection network; executing, on digital data extracted from the online connection network, one or more machine-readable queries that comprise one or more of the query string and the context data, to produce a set of candidate entities; outputting at least part of the set of candidate entities as a suggestion list that the search interface may display in association with the input field to facilitate query formulation via the input field.

Abstract: A system and method for analyzing network service usage is provided. In example embodiments, model users of the network service are identified using a number of target events and in-scope events. Users of the network service generate interaction data by interacting with the network service. The interaction data of the identified model users can be compared to the interaction data of the non-model users without exposing the interaction data.