A machine learning (ML) module that analyzes multiple language-influencing factors to correct textual content in a more meaningful and efficient manner. In the context of product support, the ML module considers the factors such as a customer's persona that shapes the words a customer chooses while speaking/writing to a customer support agent, current social trends that create new words in the social media/social platforms related to the customer's support issue, and the device used by the customer to input the textual content because different words may be input by the customer when using a smart phone versus a desktop/laptop personal computer with a traditional keyboard. The ML module also may analyze the agent's persona to modify agent's response to the customer because the agent's persona can influence the content of the agent's text. The ML module automatically corrects textual content in real-time before it is sent to the relevant recipient.

The technology disclosed proposes using a combination of computationally cheap, less-accurate bag of words (BoW) model and computationally expensive, more-accurate long short-term memory (LSTM) model to perform natural processing tasks such as sentiment analysis. The use of cheap, less-accurate BoW model is referred to herein as “skimming”. The use of expensive, more-accurate LSTM model is referred to herein as “reading”. The technology disclosed presents a probability-based guider (PBG). PBG combines the use of BoW model and the LSTM model. PBG uses a probability thresholding strategy to determine, based on the results of the BoW model, whether to invoke the LSTM model for reliably classifying a sentence as positive or negative. The technology disclosed also presents a deep neural network-based decision network (DDN) that is trained to learn the relationship between the BoW model and the LSTM model and to invoke only one of the two models.

An example method for generating a regular expression includes: acquiring a preset character string; acquiring a to-be-collected character string in the preset character string in response to a trigger instruction; recognizing a character string before the to-be-collected character string from the preset character string, the character string before the to-be-collected character string being used as a first character string; recognizing a character string after the to-be-collected character string from the preset character string, the character string after the to-be-collected character string being used as a second character string; and generating a regular expression of the to-be-collected character string by a first preset rule according to character features of the to-be-collected character string, the first character string and the second character string. The techniques of the present disclosure generate the regular expression of the character string needed by a user.

Cloud-based methods and systems for content sharing are disclosed. In some embodiments, the systems may include one or more processors configured to: receive, from a client device, an instruction for sharing a designated digital asset; retrieve the designated digital asset from a storage device; determine provenance of the designated digital asset based on metadata of the designated digital asset; generate authentication information based on the provenance of the designated digital asset; identify sensitive information in the designated digital asset; generate a redacted version of the designated digital asset by modifying content of the designated digital asset to alter the identified sensitive information; and provide the redacted version and authentication information of the designated digital asset to a recipient of the designated digital asset.

In an approach, a processor trains a model, via a reinforcement learning process, to produce a first action function for relating states of a natural language based response environment to actions applicable to the natural language based response environment. A processor retrains the model, via the reinforcement learning process, to produce a second action function, including iterations of: applying the first action function to a current state representation of the natural language based response environment to obtain a ground-truth action representation, emphasizing a word of the current state representation based on relevancy to the ground-truth action representation to obtain a modified state representation, applying a model to the modified state representation to obtain an untrained action representation, and submitting the untrained action representation to a natural language based response environment to obtain a subsequent state representation, where the subsequent state representation becomes the current state representation for a subsequent iteration.

Embodiments disclosed herein provide for dynamically modifying image and/or video content based on user preferences. Embodiments provide a server, a database containing user preference information and one or more content templates, and a content generator comprising a plurality of generative adversarial networks, wherein each generative adversarial network is associated with a corresponding style transfer, and the content generator is configured to apply one or more style transfers to a base image to convert the base image into a desired style. Once created, the customized content is sent to a client device to for display.

Embodiments disclosed herein provide for dynamically modifying image and/or video content based on user preferences. Embodiments provide a server, a database containing user preference information and one or more content templates, and a content generator comprising a plurality of generative adversarial networks, wherein each generative adversarial network is associated with a corresponding style transfer, and the content generator is configured to apply one or more style transfers to a base image to convert the base image into a desired style. Once created, the customized content is sent to a client device to for display.

An electronic data interchange (EDI) management system may comprise a memory for storing EDI document data and a machine learning prediction model representing element information of EDI documents of a first type and a corresponding status. A processor can be configured to extract elements from an EDI document, create a document record for the EDI document, the document record comprising elements extracted from the EDI document; determine a first status for the EDI document of the first type by processing the extracted elements using a machine learning model, the machine learning model trained on a training set of elements to classify documents according to a plurality of statuses, and add the first status to the document record for the EDI document of the first type, the first status accessible to a client computer via a presentation layer.

An electronic data interchange (EDI) management system may comprise a memory for storing EDI document data and a machine learning prediction model representing element information of EDI documents of a first type and a corresponding status. A processor can be configured to extract elements from an EDI document, create a document record for the EDI document, the document record comprising elements extracted from the EDI document; determine a first status for the EDI document of the first type by processing the extracted elements using a machine learning model, the machine learning model trained on a training set of elements to classify documents according to a plurality of statuses, and add the first status to the document record for the EDI document of the first type, the first status accessible to a client computer via a presentation layer.

A semantic document generation system is described. The semantic document is composed of document details, people and meta-data. The semantic document is self-aware of the information it contains. The semantic document's structure and terms are governed by legal, logical and party related rules. A semantic contract can be created from a semantic document generation system. The semantic document generation system receives an indication of a type of a document to be generated and plurality of terms for the document from a plurality of sources. The terms are converted into triples. A plurality of rules governing the terms of the document is applied to the triples to generate a knowledge graph and determine whether terms from the different parties are compatible. The terms are determined to be compatible in a case where the plurality of rules governing terms of the document is satisfied. If at least one set of terms is non-compatible, the system reconciles the non-compatible terms in the generated knowledge graph until all the terms are compatible, and generates the document based at least on the reconciled knowledge graph.