Systems and methods are provided to determine consensus values for duplicate fields in a document or form.

System and method for relation extraction using adaptive thresholding and localized context pooling (ATLOP). The system includes a computing device, the computing device has a processer and a storage device storing computer executable code. The computer executable code is configured to provide a document; embed entities in the document into embedding vectors; and predict relations between a pair of entities in the document using their embedding vectors. The relation prediction is performed based on an improved language model. Each relation has an adaptive threshold, and the relation between the pair of entities is determined to exist when a logit of the relation between the pair of entities is greater than a logit function of the corresponding adaptive threshold.

Techniques and systems are provided for predicting answers in response to one or more input queries. For instance, text from a corpus of text can be processed by a reader to generate one or multiple question and answer spaces. A question and answer space can include answerable questions and the answers associated with the questions (referred to as “question and answer pairs”). A query defining a question can be received (e.g., from a user input device) and processed by a retriever portion of the system. The retriever portion of the system can retrieve an answer to the question from the one or more pre-constructed question and answer spaces, and/or can determine an answer by comparing one or more answers retrieved from the one or more pre-constructed question and answer spaces to an answer generated by a retriever-reader system.

A method of generating a question-answer learning model through adversarial learning may include: sampling a latent variable based on constraints in an input passage; generating an answer based on the latent variable; generating a question based on the answer; and machine-learning the question-answer learning model using a dataset of the generated question and answer, wherein the constraints are controlled so that the latent variable is present in a data manifold while increasing a loss of the question-answer learning model.

Disclosed herein is a search method performed by a server, including: receiving a user question from a user terminal; generating a user question vector for the user question; selecting similar question candidates based on a similarity to the user question vector; generating an answer to the user question based on the similar question candidates; and transmitting the answer to the user question to the user terminal.

The disclosure provides a method for training a semantic retrieval network, an electronic device and a storage medium. The method includes: obtaining a training sample including a search term and n candidate files corresponding to the search term, where n is an integer greater than 1; inputting the training sample into the ranking model, to obtain n first correlation degrees output by the ranking model, in which each first correlation degree represents a correlation between a candidate document and the search term; inputting the training sample into the semantic retrieval model, to obtain n second correlation degrees output by the semantic retrieval model, wherein each second correlation degree represents a correlation between a candidate document and the search term; and training the semantic retrieval model and the ranking model jointly based on the n first correlation degrees and the n second correlation degrees.

A method and an apparatus for constructing an object relationship network and an electronic device are provided by the present disclosure, relating to the field of artificial intelligence technologies, such as deep neural networks, deep learning, etc. A specific implementation solution is: extracting keywords in respective text contents corresponding to a plurality of objects to obtain keywords corresponding to respective objects; and according to the keywords corresponding to the objects, a similarity between the plurality of objects is determined; and then according to the similarity between the plurality of objects, an object relationship network between the plurality of objects is constructed. Since the object relationship network constructed by means of the similarity between the plurality of objects can accurately describe a closeness degree of a relationship between the objects, thus, the plurality of objects can be managed effectively by means of the constructed object relationship network.

A system to calculate a reconfigured confidence score is configured to receive a text, a plurality of labels, and a plurality of confidence scores from a plurality of models and assign a weightage to the inputs received from the plurality of models. The system is configured to select a first text with a first label and retrieve a second text, a third text, and a second label. The system is further configured to generate a first, second and third output confidence score for the first text, second text and third text, and corresponding labels. The system compares the plurality of output confidence scores and generates an output which comprises of the first text, the first label, and a final confidence score, wherein the final confidence score is one among the first, second and third output confidence scores.

Techniques and systems are described for performing semantic text searches. A semantic text-searching solution uses a machine learning system (such as a deep learning system) to determine associations between the semantic meanings of words. These associations are not limited by the spelling, syntax, grammar, or even definition of words. Instead, the associations can be based on the context in which characters, words, and/or phrases are used in relation to one another. In response to detecting a request to locate text within an electronic document associated with a keyword, the semantic text-searching solution can return strings within the document that have matching and/or related semantic meanings or contexts, in addition to exact matches (e.g., string matches) within the document. The semantic text-searching solution can then output an indication of the matching strings.

Provided herein are systems, methods and computer readable media for identifying content to represent web pages and creating a representative image from the content. An example method may include retrieving a web document using a uniform resource locator (URL) contained in a dequeued work item, determining, from the web document, candidate images for creation of the representative image including extracting image references, wherein the image references are extracted by identifying image tags with source attributes, values of which are URLs locating images, filtering the URLs using a blacklist of expressions designed to match the URLs of images comprising one or more predefined undesirable characteristics, and retrieving the images which do not match any of the expressions using an HTTP client, and creating the representative image, comprising at least modifying a chosen image selected from among the candidate images.