Abstract: The present invention relates generally to the technological field of search engines. More specifically, the present invention relates to detecting statistical irregularities in data and retrieving records pertaining to these irregularities. The invention represents a system and methods of retrieving records from an indexed database which provide an improvement of the technological field of search engines and information retrieval by making the detection and retrieval of records representing statistical irregularities in the database faster, more efficient, and less computationally complex.

Abstract: A system and method of identifying occurrence of a semantic variation of a phrase in a passage by at least one processor may include calculating a phrase embedding vector, representing a semantic meaning of the phrase; extracting, from a textual representation of the passage, at least one hierarchical set of nested sequences of words; for each sequence, calculating a corresponding sequence embedding vector, representing a semantic meaning of the sequence; for one or more sequence embedding vectors, calculating a corresponding vector similarity value, representing similarity of the sequence embedding vectors to the phrase embedding vector, identifying a sequence corresponding to a maximal vector similarity value of the one or more vector similarity values; and determining the identified sequence as a semantic variation of the phrase, based on the maximal vector similarity value.

Abstract: A system and method of determining a sentiment of a participant in an interaction may include: obtaining a plurality of textual segments, each representing a portion of the interaction, and labeled according to a specific participant; inferring a language model on one or more textual segments of the plurality of textual segments, to generate respective semantic embedding vectors, each representing a semantic meaning of the respective textual segment in a semantic vector space; compiling a semantic vector set that includes (i) a target semantic embedding vector, corresponding to a target textual segment of a target participant, and (ii) one or more peripheral semantic embedding vectors, respectively corresponding to one or more peripheral textual segments of the plurality of textual segments; and inferring a composite machine-learning (ML)-based model on the semantic vector set, to classify a sentiment of the target participant, as expressed in the target textual segment.

Abstract: A method in a contact center for generating insights from conversation data derived from interactions and storing the insights in an index. The method may include: determining an insight type; based on the insight type, determining inputs including a question prompt, answer prefix, and relevant portion of the conversation data; inputting the inputs into a LLM configured to receive the inputs and generate output text answering a question contained in the question prompt pursuant to an answer form suggested by the answer prefix given content contained in the relevant portion of the conversation data; generating the output text via operation of the LLM; transforming the output text of the first insight via a sentence transformer into vector embedding representative of a semantic meaning of the output text; and storing the computed vector embedding of the first insight in the index.

Abstract: A method and system for finetuning automated sentiment classification by at least one processor may include: receiving a first machine learning (ML) model M0, pretrained to perform automated sentiment classification of utterances, based on a first annotated training dataset; associating one or more instances of model M0 to one or more corresponding sites; and for one or more (e.g., each) ML model M0 instance and/or site: receiving at least one utterance via the corresponding site; obtaining at least one data element of annotated feedback, corresponding to the at least one utterance; retraining the ML model M0, to produce a second ML model M1, based on a second annotated training dataset, wherein the second annotated training dataset may include the first annotated training dataset and the at least one annotated feedback data element; and using the second ML model M1, to classify utterances according to one or more sentiment classes.

Abstract: A method for generating a hierarchical taxonomy relating to an interaction aspect from conversation data. The method includes a first process for generating insights that includes: receiving conversation data for a first interaction; determining a conversation portion relevant to the interaction aspect and a question prompt and providing them as inputs to an LLM; and generating responsive output text via the LLM as the first insight. In a second process, inputs are provided to the LLM that include the insights, a first instruction to generate category names based on the insights, and a second instruction to make a category assignment for each insight. The second process further includes receiving from the LLM the generated category names and category assignments; grouping the insights by those having the same category assignment; and generating a hierarchical taxonomy according to the groupings.

Abstract: The present invention relates generally to the technological field of search engines. More specifically, the present invention relates to detecting statistical irregularities in data and retrieving records pertaining to these irregularities. The invention represents a system and methods of retrieving records from an indexed database which provide an improvement of the technological field of search engines and information retrieval by making the detection and retrieval of records representing statistical irregularities in the database faster, more efficient, and less computationally complex.

Abstract: A method for inverse text normalization of contact center communications according to an embodiment includes performing named entity recognition on text from a contact center communication to identify one or more entities in the text, normalizing each of the identified one or more entities in the text using weighted finite-state transducers, and normalizing at least one entity identified in the text using a large language model in response to determining that the at least one entity identified in the text was unable to be normalized using the weighted finite-state transducers.

Abstract: A method in a contact center for generating insights from conversation data derived from interactions and storing the insights in an index. The method may include: determining an insight type; based on the insight type, determining inputs including a question prompt, answer prefix, and relevant portion of the conversation data; inputting the inputs into a LLM configured to receive the inputs and generate output text answering a question contained in the question prompt pursuant to an answer form suggested by the answer prefix given content contained in the relevant portion of the conversation data; generating the output text via operation of the LLM; transforming the output text of the first insight via a sentence transformer into vector embedding representative of a semantic meaning of the output text; and storing the computed vector embedding of the first insight in the index.

Abstract: A system and method of identifying occurrence of a semantic variation of a phrase in a passage by at least one processor may include calculating a phrase embedding vector, representing a semantic meaning of the phrase; extracting, from a textual representation of the passage, at least one hierarchical set of nested sequences of words; for each sequence, calculating a corresponding sequence embedding vector, representing a semantic meaning of the sequence; for one or more sequence embedding vectors, calculating a corresponding vector similarity value, representing similarity of the sequence embedding vectors to the phrase embedding vector, identifying a sequence corresponding to a maximal vector similarity value of the one or more vector similarity values; and determining the identified sequence as a semantic variation of the phrase, based on the maximal vector similarity value.

Abstract: A method for generating automatic alternative text suggestions for a speech recognition engine of a contact center system according to an embodiment includes applying a word embedding model to generate a vector representation of each unique word in a contact center communication text corpus, calculating a cosine similarity of each vector representation and each other vector representation generated by the word embedding model, discarding each calculated cosine similarity result determined to be below a predefined threshold to generate a filtered set of word pairs, calculating a Levenshtein distance between words of each word pair of the filtered set of word pairs, and generating a candidate list of alternative words for a target word based on the Levenshtein distance between the words of each word pair of the filtered set of word pairs.

Abstract: A method for inverse text normalization of contact center communications according to an embodiment includes performing named entity recognition on text from a contact center communication to identify one or more entities in the text, normalizing each of the identified one or more entities in the text using weighted finite-state transducers, and normalizing at least one entity identified in the text using a large language model in response to determining that the at least one entity identified in the text was unable to be normalized using the weighted finite-state transducers.

Abstract: A method and system for automatic topic detection in text may include receiving a text document of a corpus of documents and extracting one or more phrases from the document, based on one or more syntactic patterns. For each phrase, embodiments of the invention may: apply a word embedding neural network on one or more words of the phrase, to obtain one or more respective word embedding vectors; calculate a weighted phrase embedding vector, and compute a phrase saliency score, based on the weighted phrase embedding vector. Embodiments of the invention may subsequently produce one or more topic labels, representing one or more respective topics in the document, based on the computed phrase saliency scores, and may select one or more topic labels according to their relevance to the business domain of the corpus.

Abstract: A system and method of automatically discovering unigrams in a speech data element may include receiving a language model that includes a plurality of n-grams, where each n-gram includes one or more unigrams; applying an acoustic machine-learning (ML) model on one or more speech data elements to obtain a character distribution function; applying a greedy decoder on the character distribution function, to predict an initial corpus of unigrams; filtering out one or more unigrams of the initial corpus to obtain a corpus of candidate unigrams, where the candidate unigrams are not included in the language model; analyzing the one or more first speech data elements, to extract at least one n-gram that comprises a candidate unigram; and updating the language model to include the extracted at least one n-gram.

Abstract: A system and method of automatically discovering unigrams in a speech data element may include receiving a language model that includes a plurality of n-grams, where each n-gram includes one or more unigrams; applying an acoustic machine-learning (ML) model on one or more speech data elements to obtain a character distribution function; applying a greedy decoder on the character distribution function, to predict an initial corpus of unigrams; filtering out one or more unigrams of the initial corpus to obtain a corpus of candidate unigrams, where the candidate unigrams are not included in the language model; analyzing the one or more first speech data elements, to extract at least one n-gram that comprises a candidate unigram; and updating the language model to include the extracted at least one n-gram.

Abstract: A first text corpus comprising punctuated and capitalized text is received. The words in the first text corpus are then annotated with a set of labels indicating a punctuation and a capitalization of each word. At an initial training stage, a machine learning model is trained on a first training set using the annotated words from the first text corpus and the labels. A second text corpus is received representing conversational speech. The words in the second text corpus are then annotated with the set of labels. In a re-training stage, the machine learning model is re-trained on a second training set comprising the annotated words from the second text corpus, and the labels. At an inference stage, the trained machine learning model is applied to a target set of words representing conversational speech to predict a punctuation and capitalization of each word in the target set.

Abstract: A method and system for automatic topic detection in text may include receiving a text document of a corpus of documents and extracting one or more phrases from the document, based on one or more syntactic patterns. For each phrase, embodiments of the invention may: apply a word embedding neural network on one or more words of the phrase, to obtain one or more respective word embedding vectors; calculate a weighted phrase embedding vector, and compute a phrase saliency score, based on the weighted phrase embedding vector. Embodiments of the invention may subsequently produce one or more topic labels, representing one or more respective topics in the document, based on the computed phrase saliency scores, and may select one or more topic labels according to their relevance to the business domain of the corpus.

Abstract: A method and system for finetuning automated sentiment classification by at least one processor may include: receiving a first machine learning (ML) model M0, pretrained to perform automated sentiment classification of utterances, based on a first annotated training dataset; associating one or more instances of model M0 to one or more corresponding sites; and for one or more (e.g., each) ML model M0 instance and/or site: receiving at least one utterance via the corresponding site; obtaining at least one data element of annotated feedback, corresponding to the at least one utterance; retraining the ML model M0, to produce a second ML model Mi, based on a second annotated training dataset, wherein the second annotated training dataset may include the first annotated training dataset and the at least one annotated feedback data element; and using the second ML model Mi, to classify utterances according to one or more sentiment classes.

Abstract: A method comprising: receiving a first text corpus comprising punctuated and capitalized text; annotating words in said first text corpus with a set of labels indicating a punctuation and a capitalization of each word; at an initial training stage, training a machine learning model on a first training set comprising: (i) said annotated words in said first text corpus, and (ii) said labels; receiving a second text corpus representing conversational speech; annotating words in said second text corpus with said set of labels; at a re-training stage, re-training said machine learning model on a second training set comprising: (iii) said annotated words in said second text corpus, and (iv) said labels; and at an inference stage, applying said trained machine learning model to a target set of words representing conversational speech, to predict a punctuation and capitalization of each word in said target set.

Abstract: A method comprising: receiving a plurality of audio segments comprising a speech signal, wherein said audio segments represent a plurality of verbal interactions; receiving labels associated with an emotional state expressed in each of said audio segments; dividing each of said audio segments into a plurality of frames, based on a specified frame duration; extracting a plurality of acoustic features from each of said frames; computing statistics over said acoustic features with respect to sequences of frames representing phoneme boundaries in said audio segments; at a training stage, training a machine learning model on a training set comprising: said statistics associated with said audio segments, and said labels; and at an inference stage, applying said trained model to one or more target audio segments comprising a speech signal, to detect an emotional state expressed in said target audio segments.