Abstract: The present invention relates generally to identifying fraudulent businesses and business listings. More specifically, the invention relates to determining a “surprisingness” value for a particular combination of words in a business title based on the likelihood that the combination has appeared in legitimate business titles. The value may be used to determine whether the business or business listing is legitimate or fraudulent. For example, third party hijackers may “keyword-stuff” business titles or attempt to include words associated with prominent businesses in a title of a less prominent business associated with the third party in order to have the less prominent business displayed more often in search results for the prominent business. For example, if a business title has too many surprising word combinations or a particular combination is highly unlikely, the business listing is likely to be fraudulent or “keyword-stuffed” and may be withheld, excluded, removed from search results.

Abstract: The present invention relates generally to identifying fraudulent businesses and business listings. More specifically, the invention relates to determining a “surprisingness” value for a particular combination of words in a business title based on the likelihood that the combination has appeared in legitimate business titles. The value may be used to determine whether the business or business listing is legitimate or fraudulent. For example, third party hijackers may “keyword-stuff” business titles or attempt to include words associated with prominent businesses in a title of a less prominent business associated with the third party in order to have the less prominent business displayed more often in search results for the prominent business. For example, if a business title has too many surprising word combinations or a particular combination is highly unlikely, the business listing is likely to be fraudulent or “keyword-stuffed” and may be withheld, excluded, removed from search results.

Abstract: The present invention provides a method for incorporating features from heterogeneous auxiliary datasets into input text data for use in classification. Heterogeneous auxiliary datasets, such as labeled datasets and unlabeled datasets, are accessed after receiving input text data. Features are extracted from each of the heterogeneous auxiliary datasets. The features are combined with the input text data to generate a set of features which may potentially be used to classify the input text data. Classification features are then extracted from the set of features and used to classify the input text data. In one embodiment, the classification features are extracted by calculating a mutual information value associated with each feature in the set of features and identifying features having a mutual information value exceeding a threshold value.

Abstract: Mechanisms are disclosed for incorporating prototype information into probabilistic models for automated information processing, mining, and knowledge discovery. Examples of these models include Hidden Markov Models (HMMs), Latent Dirichlet Allocation (LDA) models, and the like. The prototype information injects prior knowledge to such models, thereby rendering them more accurate, effective, and efficient. For instance, in the context of automated word labeling, additional knowledge is encoded into the models by providing a small set of prototypical words for each possible label. The net result is that words in a given corpus are labeled and are therefore in condition to be summarized, identified, classified, clustered, and the like.

Abstract: The present invention provides a method for incorporating features from heterogeneous auxiliary datasets into input text data for use in classification, a plurality of heterogeneous auxiliary datasets, such as labeled datasets and unlabeled datasets, are accessed after receiving input text data. A plurality of features are extracted from each of the plurality of heterogeneous auxiliary datasets. The plurality of features are combined with the input text data to generate a set of features which may potentially be used to classify the input text data. Classification features are then extracted from the set of features and used to classify the input text data. In one embodiment, the classification features are extracted by calculating a mutual information value associated with each feature in the set of features and identifying features having a mutual information value exceeding a threshold value.

Abstract: Mechanisms are disclosed for incorporating prototype information into probabilistic models for automated information processing, mining, and knowledge discovery. Examples of these models include Hidden Markov Models (HMMs), Latent Dirichlet Allocation (LDA) models, and the like. The prototype information injects prior knowledge to such models, thereby rendering them more accurate, effective, and efficient. For instance, in the context of automated word labeling, additional knowledge is encoded into the models by providing a small set of prototypical words for each possible label. The net result is that words in a given corpus are labeled and are therefore in condition to be summarized, identified, classified, clustered, and the like.