Abstract: Techniques for organizing knowledge about a dataset storing data from or about multiple sources may be provided. For example, the data can be accessed from the multiple sources and categorized based on the data type. For each data type, a triple extraction technique specific to that data type may be invoked. One set of techniques can allow the extraction of triples from the data based on natural language-based rules. Another set of techniques can allow a similar extraction based on logical or structural-based rules. A triple may store a relationship between elements of the data. The extracted triples can be stored with corresponding identifiers in a list. Further, dictionaries storing associations between elements of the data and the triples can be updated. The list and the dictionaries can be used to return triples in response to a query that specifies one or more elements.

Abstract: The present disclosure relates to systems, methods, and non-transitory computer readable media for generating query results based on domain-specific dynamic word embeddings. For example, the disclosed systems can generate dynamic vector representations of words that include domain-specific embedded information. In addition, the disclosed systems can compare the dynamic vector representations with vector representations of query terms received as part of a search query. The disclosed systems can further identify one or more digital content items to provide as part of a query result that include words corresponding to the query terms based on the comparison of the vector representations. In some embodiments, the disclosed systems can also train a word embedding model to generate accurate vector representations of unique words.

Abstract: Various embodiments describe techniques for retrieving query results for natural language procedural queries. A query answering (QA) system generates a structured semantic representation of a natural language query. The structured semantic representation includes terms in the natural language query and the relationship between the terms. The QA system retrieves a set of candidate query results for the natural language query from a repository, generates a structured semantic representation for each candidate query result, and determines a match score between the natural language query and each respective candidate query result based on the similarity between the structured semantic representations for the natural language query and each respective candidate query result. A candidate query result having the highest match score is selected as the query result for the natural language query. In some embodiments, paraphrasing rules are generated from user interaction data and are used to determine the match score.

Abstract: The present disclosure relates to generating digital responses based on digital dialog states generated by a neural network having a dynamic memory network architecture. For example, in one or more embodiments, the disclosed system provides a digital dialog having one or more segments to a dialog state tracking neural network having a dynamic memory network architecture that includes a set of multiple memory slots. In some embodiments, the dialog state tracking neural network further includes update gates and reset gates used in modifying the values stored in the memory slots. For instance, the disclosed system can utilize cross-slot interaction update/reset gates to accurately generate a digital dialog state for each of the segments of digital dialog. Subsequently, the system generates a digital response for each segment of digital dialog based on the digital dialog state.

Abstract: The present disclosure relates to generating digital responses based on digital dialog states generated by a neural network having a dynamic memory network architecture. For example, in one or more embodiments, the disclosed system provides a digital dialog having one or more segments to a dialog state tracking neural network having a dynamic memory network architecture that includes a set of multiple memory slots. In some embodiments, the dialog state tracking neural network further includes update gates and reset gates used in modifying the values stored in the memory slots. For instance, the disclosed system can utilize cross-slot interaction update/reset gates to accurately generate a digital dialog state for each of the segments of digital dialog. Subsequently, the system generates a digital response for each segment of digital dialog based on the digital dialog state.

Abstract: This disclosure relates to methods, non-transitory computer readable media, and systems that can classify term sequences within a source text based on textual features analyzed by both an implicit-class-recognition model and an explicit-class-recognition model. For example, by applying machine-learning models for both implicit and explicit class recognition, the disclosed systems can determine a class corresponding to a particular term sequence within a source text and identify the particular term sequence reflecting the class. The dual-model architecture can equip the disclosed systems to apply (i) the implicit-class-recognition model to recognize implicit references to a class in source texts and (ii) the explicit-class-recognition model to recognize explicit references to the same class in source texts.

Abstract: The present disclosure relates to systems, methods, and non-transitory computer readable media for generating modified digital images based on verbal and/or gesture input by utilizing a natural language processing neural network and one or more computer vision neural networks. The disclosed systems can receive verbal input together with gesture input. The disclosed systems can further utilize a natural language processing neural network to generate a verbal command based on verbal input. The disclosed systems can select a particular computer vision neural network based on the verbal input and/or the gesture input. The disclosed systems can apply the selected computer vision neural network to identify pixels within a digital image that correspond to an object indicated by the verbal input and/or gesture input. Utilizing the identified pixels, the disclosed systems can generate a modified digital image by performing one or more editing actions indicated by the verbal input and/or gesture input.

Abstract: The present disclosure relates to systems, methods, and non-transitory computer readable media for generating modified digital images based on verbal and/or gesture input by utilizing a natural language processing neural network and one or more computer vision neural networks. The disclosed systems can receive verbal input together with gesture input. The disclosed systems can further utilize a natural language processing neural network to generate a verbal command based on verbal input. The disclosed systems can select a particular computer vision neural network based on the verbal input and/or the gesture input. The disclosed systems can apply the selected computer vision neural network to identify pixels within a digital image that correspond to an object indicated by the verbal input and/or gesture input. Utilizing the identified pixels, the disclosed systems can generate a modified digital image by performing one or more editing actions indicated by the verbal input and/or gesture input.

Abstract: Systems and methods are discussed to automatically create a domain ontology that is a combination of ontologies. Some embodiments include systems and methods for developing a combined ontology for a website that includes extracting collocations for each webpage within the website, creating first and second ontologies from the collocations, and then aggregating the ontologies into a combined ontology. Some embodiments of the invention include unique ways to calculate collocations, to develop a smaller yet meaningful document sample from a large sample, to determine webpages of interest to users interacting with a website, and to determine topics of interest of users interacting with a website. Various other embodiments of the invention are disclosed.

Abstract: The present disclosure relates to systems, methods, and non-transitory computer readable media for generating modified digital images based on verbal and/or gesture input by utilizing a natural language processing neural network and one or more computer vision neural networks. The disclosed systems can receive verbal input together with gesture input. The disclosed systems can further utilize a natural language processing neural network to generate a verbal command based on verbal input. The disclosed systems can select a particular computer vision neural network based on the verbal input and/or the gesture input. The disclosed systems can apply the selected computer vision neural network to identify pixels within a digital image that correspond to an object indicated by the verbal input and/or gesture input. Utilizing the identified pixels, the disclosed systems can generate a modified digital image by performing one or more editing actions indicated by the verbal input and/or gesture input.

Abstract: The present disclosure relates to generating digital responses based on digital dialog states generated by a neural network having a dynamic memory network architecture. For example, in one or more embodiments, the disclosed system provides a digital dialog having one or more segments to a dialog state tracking neural network having a dynamic memory network architecture that includes a set of multiple memory slots. In some embodiments, the dialog state tracking neural network further includes update gates and reset gates used in modifying the values stored in the memory slots. For instance, the disclosed system can utilize cross-slot interaction update/reset gates to accurately generate a digital dialog state for each of the segments of digital dialog. Subsequently, the system generates a digital response for each segment of digital dialog based on the digital dialog state.

Abstract: Various embodiments describe techniques for retrieving query results for natural language procedural queries. A query answering (QA) system generates a structured semantic representation of a natural language query. The structured semantic representation includes terms in the natural language query and the relationship between the terms. The QA system retrieves a set of candidate query results for the natural language query from a repository, generates a structured semantic representation for each candidate query result, and determines a match score between the natural language query and each respective candidate query result based on the similarity between the structured semantic representations for the natural language query and each respective candidate query result. A candidate query result having the highest match score is selected as the query result for the natural language query. In some embodiments, paraphrasing rules are generated from user interaction data and are used to determine the match score.

Abstract: Systems, methods, and non-transitory computer-readable media are disclosed that collect and analyze annotation performance data to generate digital annotations for evaluating and training automatic electronic document annotation models. In particular, in one or more embodiments, the disclosed systems provide electronic documents to annotators based on annotator topic preferences. The disclosed systems then identify digital annotations and annotation performance data such as a time period spent by an annotator in generating digital annotations and annotator responses to digital annotation questions. Furthermore, in one or more embodiments, the disclosed systems utilize the identified digital annotations and the annotation performance data to generate a final set of reliable digital annotations. Additionally, in one or more embodiments, the disclosed systems provide the final set of digital annotations for utilization in training a machine learning model to generate annotations for electronic documents.

Abstract: A contextual analysis engine systematically extracts, analyzes and organizes digital content stored in an electronic file such as a webpage. Content can be extracted using a text extraction module which is capable of separating the content which is to be analyzed from less meaningful content such as format specifications and programming scripts. The resulting unstructured corpus of plain text can then be passed to a text analytics module capable of generating a structured categorization of topics included within the content. This structured categorization can be organized based on a content topic ontology which may have been previously defined or which may be developed in real-time. The systems disclosed herein optionally include an input/output interface capable of managing workflows of the text extraction module and the text analytics module, administering a cache of previously generated results, and interfacing with other applications that leverage the disclosed contextual analysis services.

Abstract: Certain embodiments identify a correct structured reading-order sequence of text segments extracted from a file. A probabilistic language model is generated from a large text corpus to comprise observed word sequence patterns for a given language. The language model measures whether splicing together a first text segment with another continuation text segment results in a phrase that is more likely than a phrase resulting from splicing together the first text segment with other continuation text segments. Sets of text segments, which include a first set with a first text segment and a first continuation text segment as well as a second set with the first text segment and a second continuation text segment, are provided to the probabilistic model. A score indicative of a likelihood of the set providing a correct structured reading-order sequence is obtained for each set of text segments.

Abstract: A computer-implemented method for providing performance indicators of destination documents includes identifying a referral document to a destination document, where the referral document comprising a source of at least one visitor to the destination document. The method also includes extracting referral keywords from content of the referral document, the referral keywords corresponding to a referral context of the referral document. The method further includes determining a degree of correlation between the referral document and the destination document based on a comparison between the referral keywords and destination keywords, the destination keywords corresponding to destination context of the destination document.

Abstract: A contextual analysis engine systematically extracts, analyzes and organizes digital content stored in an electronic file such as a webpage. Content can be extracted using a text extraction module which is capable of separating the content which is to be analyzed from less meaningful content such as format specifications and programming scripts. The resulting unstructured corpus of plain text can then be passed to a text analytics module capable of generating a structured categorization of topics included within the content. This structured categorization can be organized based on a content topic ontology which may have been previously defined or which may be developed in real-time. The systems disclosed herein optionally include an input/output interface capable of managing workflows of the text extraction module and the text analytics module, administering a cache of previously generated results, and interfacing with other applications that leverage the disclosed contextual analysis services.

Abstract: Techniques are disclosed for using natural language processing techniques to define, manipulate, and interact with consumer segmentations. In such embodiments a content consumption analytics engine can be configured to receive and process a natural language segmentation query. The query may comprise, for example, a command that defines a new segmentation, a command that manipulates existing segmentations, or a command that solicits information relating to existing consumer segmentations. The query is parsed to identify individual grammatical tokens which are then correlated with specific segment token types through the use of a token repository. A custom thesaurus is used to identify synonymous terms for grammatical tokens which may not exist in the token repository. User feedback enables the custom thesaurus to learn additional synonyms for future use.

Abstract: A computer implemented method and system identifies correct structured reading-order sequence of text segments that are extracted from a file structured in a portable document format. A probabilistic language model is generated from a large text corpus to comprise observed word sequence patterns for a given language. The language model measures whether splicing together a first text segment with another continuation text segment results in a phrase that is more likely than a phrase resulting from splicing together the first text segment with other continuation text segments. Sets of text segments are provided to the probabilistic model, where the sets of text segments comprise a first set including the first text segment and a first continuation text segment. A second set includes the first text segment and a second continuation text segment. A score is obtained for each set of text segments. The score is indicative of a likelihood of the set providing a correct structured reading-order sequence.

Abstract: Techniques for generating a query statement to query a dataset may be provided. For example, the query statement can be generated from natural language input, such as a natural language utterance. To do so, the input can be analyzed to detect a sentence, identify words in the sentence, and tag the words with the corresponding word types (e.g., nouns, verbs, adjectives, etc.). Expressions using the tags can be generated. Data about the expressions can be inputted to a classifier. Based on a detected pattern associated with the expressions, the classifier can predict a structure of the query statement, such as what expressions correspond to what clauses of the query statement. Based on this prediction, words associated with the expressions can be added to the clauses to generate the query statement and accordingly query the dataset.