Abstract: Computerized systems and methods are disclosed for performing contextual classification of objects using supervised and unsupervised training. In accordance with one implementation, content reviewers may review training objects and submit supervised training data for preprocessing and analysis. The supervised training data may be preprocessed to identify key terms and phrases, such as by stemming, tokenization, or n-gram analysis, and form vectorized objects. The vectorized objects may be used to train one or more models for subsequent classification of objects. In certain implementations, preprocessing or training, among other steps, may be performed in parallel over multiple machines to improve efficiency. The disclosed systems and methods may be used in a wide variety of applications, such as article classification and content moderation.

Abstract: Computerized systems and methods are disclosed for performing contextual classification of objects using supervised and unsupervised training. In accordance with one implementation, content reviewers may review training objects and submit supervised training data for preprocessing and analysis. The supervised training data may be preprocessed to identify key terms and phrases, such as by stemming, tokenization, or n-gram analysis, and form vectorized objects. The vectorized objects may be used to train one or more models for subsequent classification of objects. In certain implementations, preprocessing or training, among other steps, may be performed in parallel over multiple machines to improve efficiency. The disclosed systems and methods may be used in a wide variety of applications, such as article classification and content moderation.

Abstract: Computerized systems and methods are disclosed for performing contextual classification of objects using supervised and unsupervised training. In accordance with one implementation, content reviewers may review training objects and submit supervised training data for preprocessing and analysis. The supervised training data may be preprocessed to identify key terms and phrases, such as by stemming, tokenization, or n-gram analysis, and form vectorized objects. The vectorized objects may be used to train one or more models for subsequent classification of objects. In certain implementations, preprocessing or training, among other steps, may be performed in parallel over multiple machines to improve efficiency. The disclosed systems and methods may be used in a wide variety of applications, such as article classification and content moderation.

Abstract: Computerized systems and methods are disclosed for performing contextual classification of objects using supervised and unsupervised training. In accordance with one implementation, content reviewers may review training objects and submit supervised training data for preprocessing and analysis. The supervised training data may be preprocessed to identify key terms and phrases, such as by stemming, tokenization, or n-gram analysis, and form vectorized objects. The vectorized objects may be used to train one or more models for subsequent classification of objects. In certain implementations, preprocessing or training, among other steps, may be performed in parallel over multiple machines to improve efficiency. The disclosed systems and methods may be used in a wide variety of applications, such as article classification and content moderation.

Abstract: Computer-implemented systems and methods are provided for presenting content recommendations. In one implementation, a method is provided for providing content recommendations to a plurality of users. The method may include generating a first plurality of content recommendations for a first user using a first plurality of methods, each of the first plurality of methods having a weight associated therewith. The method may further include selecting a second plurality of content recommendations from the first plurality of content recommendations based on the weights associated with the first plurality of methods. The method may further include providing the second plurality of content recommendations to the first user. The method may further include updating the weights associated with the first plurality of methods based on engagement by the first user with the second plurality of content recommendations.

Abstract: Systems and methods are provided for processing and displaying user-generated content. User-generated content is received from a user over a network. A conversation related to the received user-generated content is identified. A score associated with the identified conversation is calculated based on one or more parameters related to the user or the conversation. The user-generated content is provided for display in relation to the conversation, based on the calculated score associated with the conversation.

Abstract: Computerized systems and methods are disclosed for performing contextual classification of objects using supervised and unsupervised training. In accordance with one implementation, content reviewers may review training objects and submit supervised training data for preprocessing and analysis. The supervised training data may be preprocessed to identify key terms and phrases, such as by stemming, tokenization, or n-gram analysis, and form vectorized objects. The vectorized objects may be used to train one or more models for subsequent classification of objects. In certain implementations, preprocessing or training, among other steps, may be performed in parallel over multiple machines to improve efficiency. The disclosed systems and methods may be used in a wide variety of applications, such as article classification and content moderation.

Abstract: Computerized systems and methods are disclosed for analyzing search query relationships and managing electronic content. In accordance with one implementation, log data pertaining to a plurality of queries may be received over an electronic network. A click graph may be generated representing one or more relationships between the queries. Further, temporal similarities may be identified between the queries, for example, by looking at peaks in frequency of queries over time. A pair of search queries may be evaluated based on the generated click graph and the identified temporal similarities to determine whether the queries in the pair are related.

Abstract: Computerized systems and methods are disclosed for performing contextual classification of objects using supervised and unsupervised training. In accordance with one implementation, content reviewers may review training objects and submit supervised training data for preprocessing and analysis. The supervised training data may be preprocessed to identify key terms and phrases, such as by stemming, tokenization, or n-gram analysis, and form vectorized objects. The vectorized objects may be used to train one or more models for subsequent classification of objects. In certain implementations, preprocessing or training, among other steps, may be performed in parallel over multiple machines to improve efficiency. The disclosed systems and methods may be used in a wide variety of applications, such as article classification and content moderation.

Abstract: Computerized systems and methods are disclosed for analyzing search query relationships and managing electronic content. In accordance with one implementation, log data pertaining to a plurality of queries may be received over an electronic network. A click graph may be generated representing one or more relationships between the queries. Further, temporal similarities may be identified between the queries, for example, by looking at peaks in frequency of queries over time. A pair of search queries may be evaluated based on the generated click graph and the identified temporal similarities to determine whether the queries in the pair are related.

Abstract: Systems and methods are provided for processing and displaying user-generated content. User-generated content is received from a user over a network. A conversation related to the received user-generated content is identified. A score associated with the identified conversation is calculated based on one or more parameters related to the user or the conversation. The user-generated content is provided for display in relation to the conversation, based on the calculated score associated with the conversation.

Abstract: Computer-implemented systems and methods are provided for presenting content recommendations. In one implementation, a method is provided for providing content recommendations to a plurality of users. The method may include generating a first plurality of content recommendations for a first user using a first plurality of methods, each of the first plurality of methods having a weight associated therewith. The method may further include selecting a second plurality of content recommendations from the first plurality of content recommendations based on the weights associated with the first plurality of methods. The method may further include providing the second plurality of content recommendations to the first user. The method may further include updating the weights associated with the first plurality of methods based on engagement by the first user with the second plurality of content recommendations.

Abstract: Computerized systems and methods are disclosed for analyzing search query relationships and managing electronic content. In accordance with one implementation, log data pertaining to a plurality of queries may be received over an electronic network. A click graph may be generated representing one or more relationships between the queries. Further, temporal similarities may be identified between the queries, for example, by looking at peaks in frequency of queries over time. A pair of search queries may be evaluated based on the generated click graph and the identified temporal similarities to determine whether the queries in the pair are related.

Abstract: Systems and methods are disclosed for managing electronic content, such as over the Internet. One computer-implemented method for managing electronic content includes: receiving, over an electronic network, log data of activities by Internet users; filtering the log data based on at least one aspect of the activities; aggregating the filtered log data by a predetermined period of time; calculating a trend in the aggregated log data based on a change of the activities; and presenting a request to a contributor over the electronic network, the request soliciting the submission of electronic content relating to the calculated trend.