Systems and methods for determining a query for a data store are described. A natural language text may be analyzed using heuristic processing and one or more machine learning models. Query parameters may be determined from the heuristic processing and machine learning and combined to form a query in a query language. In the heuristic processing, parsing rules may be used to remove conditional terms to generate a base question. The base question may be input to the one or more machine learning model to generate a base query which may be combined with query parameters related to the conditional terms.

Systems and methods for processing natural language inputs to determine user intents using an insights repository are provided. An insights repository system is configured to build an insights repository as a data structure representing a plurality of entities and relationships among those various entities. The insights repository system may receive information from various sources via an event stream, and may process the information using event rules. Based on the application of the event rules, the system may configure an insights repository data structure representing various entities, relationships between various entities, and the strengths of relationships between various entities. After the insights repository is created, consumers may execute queries against the insights repository. Furthermore, the insights repository system may automatically query the insights repository to generate insight information to be published to an insight feed to which consumer systems may subscribe to receive automatic updates.

In some embodiments, a method includes determining a position for a search query and a position for each audience record from multiple audience records in an embedding space. The method further includes receiving multiple device records, each associated with an audience record. The method further includes determining multiple keywords, each associated with an audience record and determining a position for each keyword in the embedding space. The method further includes calculating a first distance between the position of the search query in the embedding space and the position of each audience record in the embedding space. The method further includes calculating a second distance between the position of the search query in the embedding space and the position of each keyword in the embedding space. The method further includes ranking each audience record based on the first distance and the second distance.

A service identifies a level of specificity of one or more identified entities in a user input comprising a query, within one of multiple levels of a hierarchy of a hierarchical coding system. Responsive to determining that additional levels of specificity beyond the identified level of specificity are recommended to return a minimum answer set to the query, the service returns one or more answers requesting one or more additional inputs refining the query based on one or more values identified in a next level. Responsive to determining that no additional levels of specificity beyond the identified level of specificity are recommended to return the minimum answer set to the query, the service returns an answer set comprising a selection of information for the current level of specificity from an ingested corpus of knowledge mapped to the hierarchical coding system.

Systems and methods for conversational user experiences and conversational database analysis disclosed herein improve the efficiency and accessibility of low-latency database analytics. The method may include obtaining data expressing a usage intent with respect to the low-latency database analysis system, wherein the data expressing the usage intent includes a current request string expressed in a natural language, a current context associated with the current request string, and a previously generated context associated with a previously generated resolved-request, identifying, from the current request string, a conversational phrase corresponding to a conversational phrase pattern from a defined set of conversational phrase patterns, generating a resolved-request based on the identified conversational phrase, including the resolved-request in the current context, obtaining results data responsive to the resolved-request from a distributed in-memory database, generating a response including the results data and the current context, and outputting the response.

An apparatus comprising an interface and a natural language processor. The interface receives a data retrieval request formatted in a natural language and the natural language processor processes the data retrieval request. Processing the data retrieval request includes identifying database entities, database relations, or any combination thereof based words in the data retrieval request. It can also include identifying database entity criterion, database relation criterion, or any combination thereof based on words in the data retrieval request. It also includes generating a database query based on the database entities, the database relations, the database entity criterion, the database relation criterion, or any combination thereof and causing the database query to be applied to a database. Wherein, processing the data retrieval request includes grammatically tagging the data retrieval request using part-of-speech tagging techniques, e.g. grammatical type, grammatical context, semantic, or any combination thereof, and a database ontology.

Systems, devices, and methods discussed herein are directed to generating an answer to an input query using machine reading comprehension techniques and a lattice of supported decision trees. A supported decision tree can be generated from the various decision chains (e.g., a sequence of elements comprising a premise and a decision connected by rhetorical relationships), where the nodes of the decision tree are identified from the plurality of decision chains and ordered based on a set of predefined priority rules. A lattice may include nodes that individually correspond to a respective supported decision tree. Nodes of the lattice may be identified for an input query. The passages corresponding to those nodes may be obtained and an answer for the query may be generated from the obtained passages using machine reading comprehension techniques. The generated answer may be provided in response to the query.

The present disclosure describes a system, method, and computer program for converting a natural language update instruction to a structured update database statement. In response to receiving a natural language query for a database, an NLU model is applied to the query to identify an intent and entities associated with the query. If the intent is to update a data object, the system evaluates the entities to identify update fields and update values. Update fields are matched to update values based on update parameters, operand type of the update value, and location of the update fields and values. For each update field and value pair, an update context is calculated to determine whether the update value is absolute or relative to an existing field value. An update plan is created with the update field and value pairs and corresponding update contexts, and a database update statement is generated from the update plan.

A system and method for translating questions into database queries are provided. A text to database query system receives a natural language question and a structure in a database. Question tokens are generated from the question and query tokens are generated from the structure in the database. The question tokens and query tokens are concatenated into a sentence and a sentence token is added to the sentence. A BERT network generates question hidden states for the question tokens, query hidden states for the query tokens, and a classifier hidden state for the sentence token. A translatability predictor network determines if the question is translatable or untranslatable. A decoder converts a translatable question into an executable query. A confusion span predictor network identifies a confusion span in the untranslatable question that causes the question to be untranslatable. An auto-correction module to auto-correct the tokens in the confusion span.

A computer retrieves data from a database. The computer retrieves a Machine Learning (ML) model trained to generate database queries. The computer applies the ML model to generate a primary database query based, at least in part, on a user inquiry available to the computer. The computer retrieves the primary database query, an initial set of data from a database available to the computer. The computer, in response to retrieving the initial set of data, receives feedback assessing the initial set of data. The computer, in response to receiving the feedback, applies a Natural Language Processing (NLP) model to identify query adjustment content within the feedback. The computer revises the primary database query based, at least in part, on the model adjustment content, to generate a secondary database query. The computer retrieves using the secondary database query, a secondary set of data from the database.