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.

An information search system, including: a database (12); a query sentence acceptance unit (26) that accepts a query sentence; an inputted search keyword extractor (44) that extracts an inputted search keyword from the query sentence; a shared keyword dictionary (30) in which relevant keywords are registered in association with each other; a local keyword dictionary (102) in which district keywords used in particular districts are registered; a candidate search keyword reader (32) that reads out a keyword that is relevant to the inputted search keyword; and a retrieval executor (40) that executes retrieval processing from the database using the inputted search keyword, wherein, in a case in which the inputted search keyword is not registered in the local keyword dictionary, the candidate search keyword reader refers to the shared keyword dictionary, so as to read out a keyword that is relevant to the inputted search keyword.

A business analytics conversational tool comprising: a device comprising a communication channel, a natural language processor (NLP), a fulfillment application program interface (F-API), a database application program interface (D-API), and a business management database; wherein: the NLP receives a user-input from a user through the communication channel; the NLP deduces an intent of the user-input; the NLP communicates the intent to the F-API; the F-API communicates a request for data associated with the intent to the database via the D-API; the D-API communicates the data associated with the intent to the F-API; the F-API converts the data associated with the intent to conversational form and sends the conversational form for voice output through the communication channel.

Provided is a learning assistance technology that uses a learning history to check the level of comprehension by a learner in relation to a learning target. Included are a score calculation unit that uses a first occurrence ratio α(n) of a learning target Q(n) calculated using an occurrence frequency R(n) of the learning target Q(n) in a document to be used as a basis for creating a confirmation question and a second occurrence ratio β(n) of the learning target Q(n) weighted by viewing time and calculated using the occurrence frequency R(n) of the learning target Q(n) and a viewing time for each page of the document included in a learning history to calculate one of the difference between the first occurrence ratio α(n) and the second occurrence ratio β(n), the absolute value of the difference, or the ratio as a score S(n) of the learning target Q(n), and a query generation unit that treats the learning target Q(n) corresponding to the n for which the score S(n) is maximized as a query, that is, the learning target with which to create a confirmation question for a learner.

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.

Implementations are directed to determining, based on a submitted query that is a compound query, that a set of multiple sub-queries are collectively an appropriate interpretation of the compound query. Those implementations are further directed to providing, in response to such a determination, a corresponding command for each of the sub-queries of the determined set. Each of the commands is to a corresponding agent (of one or more agents), and causes the agent to generate and provide corresponding responsive content. Those implementations are further directed to causing content to be rendered in response to the submitted query, where the content is based on the corresponding responsive content received in response to the commands.

Disclosed is a solution for diagnosing problems from logs used in an application development environment. A random sample of log statements is collected. The log statements can be completely unstructured and/or do not conform to any natural language. The log statements are tagged with predefined classifications. A natural language processing (NLP) classifier model is trained utilizing the log statements tagged with the predefined classification. New log statements can be classified into the plurality of predefined classifications utilizing the trained NLP classifier model. From the log statements thus classified, statements having a problem classification can be identified and presented through a dashboard running in a browser. Outputs from the trained NLP classifier model can be provided as input to another trained model for automatically and quickly identifying a type of problem associated with the statements, eliminating a need to manually sift through tens or hundreds of thousands of lines of logs.

A method and system may be provided for recording discussions about computer code in an integrated development environment (“IDE”). In some aspects, a communication channel is integrated with an IDE. Communications and discussions may be tracked and linked with specific code sections.

A method trains a neural network to identify an event based on discrepancies in answers to factoid questions at different times. One or more processors identify answers to a series of factoid questions. The processor(s) compare the answers from the series of factoid questions in order to determine discrepancies in the answers at different times, and then train a neural network to identify an event based on the discrepancies in the answers at the different times.

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for using frames for performing tasks. One of the methods includes receiving a first request to perform a task, the first request comprising user speech identifying the task; generating a frame associated with the task, wherein the frame comprises one or more types of values necessary to perform the task, and wherein each type of value can be satisfied by a respective value; receiving a second request to provide information related to a question, the second request comprising user speech identifying the question; providing information identifying the question to a search engine, and receiving a response identifying one or more terms; determining that at least one term can satisfy a type of value necessary to perform the task; and storing the at least one term in the frame.