Abstract: Online abstract groups, in which members aren't explicitly connected, can be automatically identified by computer-implemented methods. The methods involve harvesting records from social media and extracting content-based and structure-based features from each record. Each record includes a social-media posting and is associated with one or more entities. Each feature is stored on a data storage device and includes a computer-readable representation of an attribute of one or more records. The methods further involve grouping records into record groups according to the features of each record. Further still the methods involve calculating an n-dimensional surface representing each record group and defining an outlier as a record having feature-based distances measured from every n-dimensional surface that exceed a threshold value. Each of the n-dimensional surfaces is described by a footprint that characterizes the respective record group as an online abstract group.

Abstract: Online abstract groups, in which members aren't explicitly connected, can be automatically identified by computer-implemented methods. The methods involve harvesting records from social media and extracting content-based and structure-based features from each record. Each record includes a social-media posting and is associated with one or more entities. Each feature is stored on a data storage device and includes a computer-readable representation of an attribute of one or more records. The methods further involve grouping records into record groups according to the features of each record. Further still the methods involve calculating an n-dimensional surface representing each record group and defining an outlier as a record having feature-based distances measured from every n-dimensional surface that exceed a threshold value. Each of the n-dimensional surfaces is described by a footprint that characterizes the respective record group as an online abstract group.

Abstract: Desired content, metadata, or both can be isolated from the full content of social media websites having content-rich pages. Achieving this can include obtaining from the content-rich pages a language-independent representation having a hierarchical structure of nodes and then generating a node representation for each node. Feature vectors for the nodes are generated and a label is assigned to each node representation according to a schema. Assignment can occur by executing a trained classification algorithm on the feature vectors. The schema has schema elements and each schema element corresponds to a label. For each schema element, all node representations having matching labels are gathered and then one node representation is elected from among those with matching labels to be assigned to a schema element field in a template. The template can be applied to extract desired content, metadata, or both according to the schema from all the content-rich pages.

Abstract: Desired content, metadata, or both can be isolated from the full content of social media websites having content-rich pages. Achieving this can include obtaining from the content-rich pages a language-independent representation having a hierarchical structure of nodes and then generating a node representation for each node. Feature vectors for the nodes are generated and a label is assigned to each node representation according to a schema. Assignment can occur by executing a trained classification algorithm on the feature vectors. The schema has schema elements and each schema element corresponds to a label. For each schema element, all node representations having matching labels are gathered and then one node representation is elected from among those with matching labels to be assigned to a schema element field in a template. The template can be applied to extract desired content, metadata, or both according to the schema from all the content-rich pages.

Abstract: Hypothesis analysis methods, hypothesis analysis devices, and articles of manufacture are described according to some aspects. In one aspect, a hypothesis analysis method includes providing a hypothesis, providing an indicator which at least one of supports and refutes the hypothesis, using the indicator, associating evidence with the hypothesis, weighting the association of the evidence with the hypothesis, and using the weighting, providing information regarding the accuracy of the hypothesis.

Abstract: Methods of defining ontologies, word disambiguation methods, computer systems, and articles of manufacture are described according to some aspects. In one aspect, a word disambiguation method includes accessing textual content to be disambiguated, wherein the textual content comprises a plurality of words individually comprising a plurality of word senses, for an individual word of the textual content, identifying one of the word senses of the word as indicative of the meaning of the word in the textual content, for the individual word, selecting one of a plurality of event classes of a lexical database ontology using the identified word sense of the individual word, and for the individual word, associating the selected one of the event classes with the textual content to provide disambiguation of a meaning of the individual word in the textual content.

Abstract: Lexicon generation methods, computer implemented lexicon editing methods, lexicon generation devices, lexicon editors, and articles of manufacture are described according to some aspects. In one aspect, a lexicon generation method includes providing a seed vector indicative of occurrences of a plurality of seed terms within a plurality of text items, providing a plurality of content vectors indicative of occurrences of respective ones of a plurality of content terms within the text items, comparing individual ones of the content vectors with respect to the seed vector, and responsive to the comparing, selecting at least one of the content terms as a term of a lexicon usable in sentiment analysis of text.

Abstract: Methods, apparatus, and computer-readable media for analyzing conversational-type data by association of two or more types of extracted information in view of time are disclosed according to some aspects. In one embodiment, analysis of conversational-type data comprises identification of topical segments within the conversational-type data and linking of the topical segments with at least one other type of pertinent, extracted information. The linking can be based on a sequential order of utterances that compose, at least in part, the conversational-type data.