Microsoft Technology Licensing, LLC

Jian-Guang Lou

Dongmei Zhang

Haidong Zhang

Bo Zhang

Yan Gao

Bei Chen

He Huang

Xiao Bi Liang

Qingwei Lin

Zhitao Hou

Maria Del Mar Gines Marin

Daniel M Cheung

Yu Mao

Lei Fang

Yan Xiao

Ishita Gupta

Xiao Liang

Weiwei Cui

Implementations of the subject matter described herein relate to conversational data analysis. After a data analysis request is received from a user, heuristic information may be determined based on the data analysis request. The heuristic information mentioned here is not a result for the data analysis request but information which may be used for leading the conversation to proceed. Based on such heuristic information, the user may provide supplementary information associated with the data analysis request, for example, clarify meaning of the data analysis request, submit a relevant further analysis request, and so on. A really desired and meaningful data analysis result can be provided to the user according to the supplementary information provided by the user. Thus, data analysis will become more accurate and effective. While obtaining really helpful information, the user also gains good user experience.

Conversational data analysis

According to implementations of the subject matter described herein, a solution for semantic parsing of the natural language query is provided. In this solution, a first natural language query for a first data table is received, where the first natural language query comprises words. A first set of knowledge representations associated with the first data table are obtained and the first set of knowledge representations describe, in a structured manner, knowledge required for understanding a natural language query for the first data table. The first natural language query is converted into a first sequence of symbols. A first computer-executable query corresponding to the first natural language query is generated. According to this solution, various kinds of knowledge representations from different fields can be represented into universal structured knowledge representations, such that the knowledge representations are used to assist in understanding semantics of natural language queries and generating corresponding computer-executable queries.

Semantic parsing of natural language query

Various implementations of the present disclosure relate to text to visualization. In a method, information items are extracted from a natural language sentence. Visual elements associated with the information items are determined. A visual representation of the natural language sentence based on the visual elements is determined, the visual representation indicating information expressed by the natural language sentence.

Text to visualization

In accordance with implementations of the present disclosure, a solution for converting a natural language query is provided. In this solution, a first natural language query and a second natural language query for one or more data tables are received, wherein semantics of the second natural language query is dependent on the first natural language query. A third natural language query for one or more data tables is generated based on the first natural language query and the second natural language query, wherein semantics of the third natural language query is identical to the semantics of the second natural language query and independent of the first natural language query. In this way, this solution can convert a context-dependent natural language query into a context-independent natural language query, thereby enabling interfacing with any semantic parsers which can convert a natural language query into a computer-executable query.

Conversion of natural language query

In implementations of the subject matter described herein, a solution for query processing is provided. In this solution, data subsets are pre-stored for example in a fast access storage device for data analysis, each including data entries corresponding to one or more dimensions. If two or more data subsets are needed to cover target dimensions corresponding to query items in a received query, instead of turning to analyze a source data set that is not stored, the query is decomposed into subqueries. By means of the decomposing, the target dimension(s) corresponding to the query item(s) in each subquery can be covered by a single data subset. The data subset is analyzed for each subquery and a query result for the query is determined based on analysis results of the subqueries. In such way, the query result for the query can obtained in a fast manner from the available data subsets.

Query processing in data analysis

Growing community of inventors

Redmond, WA, United States of America

Jian-Guang Lou