Abstract: Methods and systems for receiving a plurality of documents including short text data and determining a plurality of forward similarity values based on the short text data in each of the plurality of documents, determining a plurality of reverse similarity values based on the short text data in each of the plurality of documents, generating a forward and reverse similarity matrix based on the plurality of forward similarity values and the plurality of reverse similarity values, and generating a plurality of short text similarity based clusters to group the short text data of the plurality of documents based on the forward and reverse similarity matrix.

Abstract: A method and system of language independent iterative learning mechanism for Natural Language Processing (NLP) tasks is disclosed. The method includes identifying at least one NLP feature associated with a set of words within a sentence for an NLP task. The method includes creating a pattern associated with the sentence for the NLP task, based on the at least one NLP feature associated with the set of words and the linkage relationship between each subset of two adjacent words. The method further includes computing a confidence score corresponding to the pattern, based on a comparison within a trained dataset. The method further includes assigning a pattern category to the pattern, based on the confidence score and a predefined threshold score. The method further includes executing the NLP task based on the assigned pattern category.

Abstract: The present disclosure relates to a method and system for generating question variations to user input. The method comprises receiving a user input comprising at least one sentence from a user. Further, plurality of keywords and associated plurality of features are extracted from the at least one sentence. Thereafter, a plurality of question variations is generated for the user input by using one or more subgraphs identified from a trained knowledge graph based on the plurality of keywords and the associated plurality of features. The plurality of keywords and the associated plurality of features are extracted using rich semantics processing, transformation of words from one form to another and similarity. The one or more subgraphs include an entity specific graph and an action specific graph. The system and method of the present disclosure fetches real time user input and generates probable question variations for the user input.

Abstract: The present disclosure relates to method and system for providing context-based response for a user query. The method comprises generating current graphical structure for user query received from user which comprises one or more nodes corresponding to one or more words in the user query and one or more edges connecting the one or more nodes. Each node in the current graphical structure is associated with predefined indexing identification (ID) which is compared with predefined indexing IDs of one or more predefined graphical structures to identify one or more related predefined graphical structures. The one or more related predefined graphical structures are ranked based on number of predefined indexing IDs being matched. Based on comparison, system provides response corresponding to predefined question associated with related predefined graphical structure with highest rank. Therefore, present disclosure provides an efficient mechanism for providing context-based response to user queries.

Abstract: The present disclosure relates to method and system for providing context-based response for a user query. The method comprises generating current graphical structure for user query received from user which comprises one or more nodes corresponding to one or more words in the user query and one or more edges connecting the one or more nodes. Each node in the current graphical structure is associated with predefined indexing identification (ID) which is compared with predefined indexing IDs of one or more predefined graphical structures to identify one or more related predefined graphical structures. The one or more related predefined graphical structures are ranked based on number of predefined indexing IDs being matched. Based on comparison, system provides response corresponding to predefined question associated with related predefined graphical structure with highest rank. Therefore, present disclosure provides an efficient mechanism for providing context-based response to user queries.

Abstract: A method and system of language independent iterative learning mechanism for Natural Language Processing (NLP) tasks is disclosed. The method includes identifying at least one NLP feature associated with a set of words within a sentence for an NLP task. The method includes creating a pattern associated with the sentence for the NLP task, based on the at least one NLP feature associated with the set of words and the linkage relationship between each subset of two adjacent words. The method further includes computing a confidence score corresponding to the pattern, based on a comparison within a trained dataset. The method further includes assigning a pattern category to the pattern, based on the confidence score and a predefined threshold score. The method further includes executing the NLP task based on the assigned pattern category.

Abstract: The present disclosure discloses method and response generation system for generating dialogue responses for users in real-time. The response generation system receives pair of utterance from one or more sources, where pair of utterance comprise conversation between user and computing device, identifies one or more labels from utterance of pair of utterance that corresponds to user, where one or more labels exhibit relationship. One or more intent corresponding to each of identified one or more labels is determined, where one or more intent is restricted according to current context of utterance of user. A polarity of each of one or more intent is determined based on pre-defined scale comprising left bound and right bound and one or more responses for user are generated until determined polarity for each of labels from identified one or more labels is within pre-defined distance of left bound or right bound of pre-defined scale.

Abstract: The present disclosure relates to a method and system for generating question variations to user input. The method comprises receiving a user input comprising at least one sentence from a user. Further, plurality of keywords and associated plurality of features are extracted from the at least one sentence. Thereafter, a plurality of question variations is generated for the user input by using one or more subgraphs identified from a trained knowledge graph based on the plurality of keywords and the associated plurality of features. The plurality of keywords and the associated plurality of features are extracted using rich semantics processing, transformation of words from one form to another and similarity. The one or more subgraphs include an entity specific graph and an action specific graph. The system and method of the present disclosure fetches real time user input and generates probable question variations for the user input.

Abstract: A method and device for n-gram identification and extraction is disclosed. The method includes identifying at least one n-gram from a sentence inputted by a user based on a confidence score associated with each of the at least one n-gram. The method further includes determining a direction context entropy coefficient for each of the at least one n-gram. The method includes iteratively expanding one or more of the at least one n-gram by the smallest n-gram unit at each iteration in a predefined direction in the sentence to generate at least one expanded n-gram, based on an associated direction context entropy coefficient. The method further includes extracting at each expanding iteration one or more of the at least one expanded n-gram based on an associated confidence score. The method includes grouping semantically linked n-grams from the one or more of the at least one expanded n-gram.

Abstract: This disclosure relates generally to natural language processing, and more particularly to system and method for generating named entities. In one embodiment, a method is provided for generating named entities. The method includes extracting a plurality of named entities in a primary language from a plurality of digital content in the primary language, transliterating each of the plurality of named entities in the primary language to a set of possible named entities in a secondary language, determining a correct named entity in the secondary language from among the set of possible named entities in the secondary language, and generating a named entity in a subsequent secondary language corresponding to the correct named entity in the secondary language. It should be noted that the plurality of named entities in the primary language are named entities in the subsequent secondary language, and the subsequent secondary language is related to the secondary language.

Abstract: A method and device for extracting images from PDF documents are disclosed. The method includes performing a text recognition process on a PDF document that includes one or more images. The text recognition process replaces the one or more images with a plurality of contiguous newlines. The method further includes storing a location of each of the one or more images within the PDF document based on occurrence of the plurality of contiguous newlines within the PDF document. The method includes converting each page of the PDF document to an image format in order to generate an image document corresponding to the PDF document. The method further includes extracting each of the one or more images from the image document based on the location stored for each of the one or more images within the PDF document.

Abstract: The present disclosure discloses method and response generation system for generating dialogue responses for users in real-time. The response generation system receives pair of utterance from one or more sources, where pair of utterance comprise conversation between user and computing device, identifies one or more labels from utterance of pair of utterance that corresponds to user, where one or more labels exhibit relationship. One or more intent corresponding to each of identified one or more labels is determined, where one or more intent is restricted according to current context of utterance of user. A polarity of each of one or more intent is determined based on pre-defined scale comprising left bound and right bound and one or more responses for user are generated until determined polarity for each of labels from identified one or more labels is within pre-defined distance of left bound or right bound of pre-defined scale.

Abstract: A method and device for n-gram identification and extraction is disclosed. The method includes identifying at least one n-gram from a sentence inputted by a user based on a confidence score associated with each of the at least one n-gram. The method further includes determining a direction context entropy coefficient for each of the at least one n-gram. The method includes iteratively expanding one or more of the at least one n-gram by the smallest n-gram unit at each iteration in a predefined direction in the sentence to generate at least one expanded n-gram, based on an associated direction context entropy coefficient. The method further includes extracting at each expanding iteration one or more of the at least one expanded n-gram based on an associated confidence score. The method includes grouping semantically linked n-grams from the one or more of the at least one expanded n-gram.

Abstract: A method and device for extracting images from PDF documents are disclosed. The method includes performing a text recognition process on a PDF document that includes one or more images. The text recognition process replaces the one or more images with a plurality of contiguous newlines. The method further includes storing a location of each of the one or more images within the PDF document based on occurrence of the plurality of contiguous newlines within the PDF document. The method includes converting each page of the PDF document to an image format in order to generate an image document corresponding to the PDF document. The method further includes extracting each of the one or more images from the image document based on the location stored for each of the one or more images within the PDF document.

Abstract: This disclosure relates generally to natural language processing, and more particularly to system and method for generating named entities. In one embodiment, a method is provided for generating named entities. The method includes extracting a plurality of named entities in a primary language from a plurality of digital content in the primary language, transliterating each of the plurality of named entities in the primary language to a set of possible named entities in a secondary language, determining a correct named entity in the secondary language from among the set of possible named entities in the secondary language, and generating a named entity in a subsequent secondary language corresponding to the correct named entity in the secondary language. It should be noted that the plurality of named entities in the primary language are named entities in the subsequent secondary language, and the subsequent secondary language is related to the secondary language.