Abstract: Disclosed in some examples are methods, systems, and machine-readable mediums which provide for a personalized expertise searching. When a user of the social networking service enters a search query, the system determines if the user is searching for members who possess a particular skill. If the user is searching for members who possess a particular skill, the search results are post-processed by personalizing the search results using one or more machine-learning models which utilize one or more observed features about the user that enters the query, the skills of the members of the social networking service, and the query itself. In some examples, the system may utilize multiple machine-learning models in multiple passes to fine tune the relevance of the search results and to ensure that the post-processing returns search results in a timely manner.

Abstract: Disclosed in some examples are methods, systems, and machine-readable mediums which provide for a personalized expertise searching. When a user of the social networking service enters a search query, the system determines if the user is searching for members who possess a particular skill. If the user is searching for members who possess a particular skill, the search results are post-processed by personalizing the search results using one or more machine-learning models which utilize one or more observed features about the user that enters the query, the skills of the members of the social networking service, and the query itself. In some examples, the system may utilize multiple machine-learning models in multiple passes to fine tune the relevance of the search results and to ensure that the post-processing returns search results in a timely manner.

Abstract: A system and method for building a language model for a translation system are provided. The method includes providing a first relative ranking of first and second translations in a target language of a same source string in a source language, determining a second relative ranking of the first and second translations using weights of a language model, the language model including a weight for each of a set of n-gram features, and comparing the first and second relative rankings to determine whether they are in agreement. The method further includes, when the rankings are not in agreement, updating one or more of the weights in the language model as a function of a measure of confidence in the weight, the confidence being a function of previous observations of the n-gram feature in the method.

Abstract: A classification method includes constructing queries from category descriptors representing categories of a taxonomy of hierarchically organized categories. The query constructed for a category c includes a query component based on descriptors of the category c and at least one query component based on descriptors of an ancestor or descendant category of the category c. A documents database is queried using the constructed queries to retrieve pseudo-relevant documents. Language models for the categories of the taxonomy are extracted from the pseudo-relevant documents by inferring a hierarchical topic model representing the taxonomy. An input document is classified by optimizing mixture weights of a weighted combination of categories of the hierarchical topic model respective to the input document.

Abstract: A system and method for building a language model for a translation system are provided. The method includes providing a first relative ranking of first and second translations in a target language of a same source string in a source language, determining a second relative ranking of the first and second translations using weights of a language model, the language model including a weight for each of a set of n-gram features, and comparing the first and second relative rankings to determine whether they are in agreement. The method further includes, when the rankings are not in agreement, updating one or more of the weights in the language model as a function of a measure of confidence in the weight, the confidence being a function of previous observations of the n-gram feature in the method.

Abstract: A classification method includes constructing queries from category descriptors representing categories of a taxonomy of hierarchically organized categories. The query constructed for a category c includes a query component based on descriptors of the category c and at least one query component based on descriptors of an ancestor or descendant category of the category c. A documents database is queried using the constructed queries to retrieve pseudo-relevant documents. Language models for the categories of the taxonomy are extracted from the pseudo-relevant documents by inferring a hierarchical topic model representing the taxonomy. An input document is classified by optimizing mixture weights of a weighted combination of categories of the hierarchical topic model respective to the input document.