Microsoft Technology Licensing, LLC

Thoughtspot, Inc.

Sanjay Agrawal

Surajit J Chaudhuri

Vivek Ravindranath Narasayya

Venkatesh Ganti

Amit Prakash

Kaushik Chakrabarti

Gunjan Jha

Antony Chuxiao Chen

Ashok Anand

Bhanu Prakash

Guozhang Wang

Atul Adya

Ashish Gupta

Beverly Yang

Shashank Gupta

Ajeet Singh

Manoj Krishna Ghosh

Pavan Ram Piratla

Abhishek Rai

Jasmeet Singh Jaggi

Anish Das Sarma

Abhinandan Sujit Das

Priyendra Singh Deshwal

Guatam Das

Joy Dutta

Vijay Krishnan Ganesan

Vibhor Nanavati

Stephane Antonin Kiss

Yuri Siradeghyan

Lik Han Chu

Pradeep Dorairaj

A request for database results is received from a query coordinator at a database instance of a distributed database. The request includes a query execution instruction of a query plan and an indication of override instructions corresponding to the query execution instruction. The override instructions are such that they do not modify the query plan. The database instance includes the override instructions in a set of high-level language query instructions. The database instance performs just-in-time compiling of the set of high-level language query instructions to obtain a machine language query for performing the query execution instruction of the query plan. The database instance executes the machine language query to obtain the database results. The database instance then transmits the database results to the query coordinator.

Just-in-time injection in a distributed database

Injecting override instructions associated with query executions instructions performed on a distributed database includes receiving a data-query; generating, by a first database instance, a query plan that includes a first query execution instruction for transmission to a second database instance; transmitting, by the first database instance, a request for partial results to the second database instance, where the request includes the first query execution instruction and an indication of override instructions corresponding to the first query execution instruction; responsive to a determination that the request includes the indication, including, by the second database instance, the override instructions in a set of high-level language query instructions; obtaining, by the second database instance, a machine language query based on the set; executing, at the second database instance, the machine language query to obtain the partial results; and transmitting, by the second database instance, the partial results to the first database instance.

Automatic database analysis includes identifying a current context for accessing data from a low-latency database and generating an exploration query based on the current context, which includes identifying a column from the low-latency database as a column of utility in response to determining that a probabilistic utility for the column satisfies a defined utility criterion. The current context includes a requested result set satisfying a requested search criterion, and the probabilistic utility is based on the current context. The analysis includes generating an exploration result set based on the exploration query, generating insights based on the exploration result set, ranking the insights, and outputting at least one insight based on the ranking.

Automatic database insight analysis

Low-latency autonomous-analysis includes obtaining data expressing a usage intent with respect to a low-latency database analysis system that intent omits data corresponding to user input expressly requesting low-latency autonomous-analysis, obtaining requested results data based on the data expressing the usage intent, outputting requested visualization data representing at least a portion of the requested results data for presentation to a user, and, in response to outputting the requested visualization data, obtaining low-latency autonomous-analysis data by performing low-latency autonomous-analysis based on the data expressing the usage intent by identifying an autonomous-analysis predicate based on the requested visualization data, obtaining a defined autonomous-analysis latency constraint, obtaining the low-latency autonomous-analysis data based on the autonomous-analysis predicate in accordance with the defined autonomous-analysis latency constraint, such that the low-latency autonomous-analysis data differs from the requested results data, and outputting at least a portion of the low-latency autonomous-analysis data for presentation to a user.

Low-latency predictive database analysis

The claimed subject matter provides a system and method for automatic allocation of tenants to database services. An exemplary method comprises determining a query plan based on a workload for each tenant. A distance metric is created based on the query plan for each tenant, and tenants are clustered based on the distance metric into a predetermined number of clusters which is equal to the predetermined number of databases.

Allocation of tenants to database services

A method to automatically allocate tenants to database services includes determining a query plan based on a workload for each tenant. A distance metric is created based on the query plan for each tenant, and tenants are clustered based on the distance metric into a predetermined number of clusters which is equal to the predetermined number of databases.

A set of documents is filtered for entity extraction. A list of entity strings is received. A set of token sets that covers the entity strings in the list is determined. An inverted index generated on a first set of documents is queried using the set of token sets to determine a set of document identifiers for a subset of the documents in the first set. A second set of documents identified by the set of document identifiers is retrieved from the first set of documents. The second set of documents is filtered to include one or more documents of the second set that each includes a match with at least one entity string of the list of entity strings. Entity recognition may be performed on the filtered second set of documents.

Scalable lookup-driven entity extraction from indexed document collections

An information retrieval system converts unstructured ad-hoc search queries into structured search instructions that retrieve data in a structured relational database or an unstructured database. Data from the database is uploaded into a distributed in-memory database system. Tokens are automatically generated based on attributes, measures, and other metadata extracted from the relational database. The tokens are then compared with the non-structured ad-hoc user search queries. The information retrieval system uses the tokens to identify or predict what structured data is associated with user search queries. The tokens guide the user through a set of search terms that the system then uses to generate the structured query instructions. The structured query instructions retrieve specific data and answers from in the database system.

Information retrieval system

Described is using context information obtained from entity mentions in likely relevant documents to extract entity mentions from documents that are ambiguous with respect to their relevance to a domain. A list of entities is input into an entity extraction mechanism, which processes a large collection of documents to determine data (counts) corresponding to frequency of entity mentions. Infrequently mentioned entities are specific entities, while frequently mentioned entities are non-specific (generic or ambiguous) entities. The context surrounding mentions of the specific entities is processed to obtain interesting context terms (words, phrases or both) for the domain. The interesting context terms are then compared against the contexts of non-specific entity mentions to determine whether each non-specific entity mention is relevant to the domain. A result set containing only relevant documents or relevant mentions collection is output.

Using context to extract entities from a document collection

This patent application relates to taxonomy editing. One implementation involves a taxonomy editor configured to generate a visual representation of a taxonomy associated with a set of scientific papers. The taxonomy editor includes a properties module configured to identify properties relating to an individual node of the taxonomy and a statistics module configured to determine trends relating to the individual node. The taxonomy editor further includes a similarity module configured to evaluate keyword similarity relative to individual scientific papers associated with the individual node. The taxonomy editor also includes a suggestion module configured to utilize the properties, the trends and the keyword similarity to identify potential modifications to the taxonomy. The taxonomy editor is further configured to present at least some of the potential modifications, the properties, the trends, and the keyword similarity concurrently with the visual representation of the taxonomy.

Taxonomy editor

Techniques are described for incorporating different types of input available to administrators responsible for assigning reviewers (persons) to objects to be reviewed. A plurality of types of assignment ranking algorithms is provided. Each ranking algorithm can output a different set of candidate assignments (reviewer-review object pairs) with corresponding ranks/weights (measures of how well object fit reviewers). Some ranking algorithms may automatically generate assignments by using information about review objects and reviewers to find matches therebetween. Other algorithms might simply rank or weigh suggested matches (e.g., from manual input). The ranked/weighted assignments from the various algorithms are combined to produce a final result set of suggested assignments of reviewers. In one embodiment, a bidding algorithm, referral algorithm, and a relevance algorithm may be used, and intermediate results therefrom are combined for a final result.

Matching reviewers to review objects

Architecture for finding related entities for web search queries. An extraction component takes a document as input and outputs all the mentions (or occurrences) of named entities such as names of people, organizations, locations, and products in the document, as well as entity metadata. An indexing component takes a document identifier (docID) and the set of mentions of named entities and, stores and indexes the information for retrieval. A document-based search component takes a keyword query and returns the docIDs of the top documents matching with the query. A retrieval component takes a docID as input, accesses the information stored by the indexing component and returns the set of mentions of named entities in the document. This information is then passed to an entity scoring and thresholding component that computes an aggregate score of each entity and selects the entities to return to the user.

Finding related entity results for search queries

The subject disclosure pertains to exploitation of sequence information in a workload for the purpose of performance tuning. An optimal or exhaustive approach to tune sequences by mapping to a shortest path problem is provided as well as powerful techniques that result in optimal or nearly optimal solutions. Also disclosed is greedy approach that is much more efficient than the optimal approach and yet also generates solutions of comparable quality. Further yet, systems and methods are provided that facilitate extraction of sequence information and implementation of recommendations.

Growing community of inventors

Sammamish, WA, United States of America

Sanjay Agrawal