Microsoft Technology Licensing, LLC

Shi Han

Dongmei Zhang

Jian Wang

Yu Zou

Ming Chang

Rui Ding

Yingnong Dang

Zhouyu Fu

Haoyu Dong

Song Ge

Haidong Zhang

Xinjian Chen

Peng Liu

Chunhui Zhang

Lei Ma

Jennifer Michelstein Halberstam

Frank Soong

Yi-Jian Wu

Xiao Lyu

Zhengyou Zhang

Hang Li

Bin Zhao

Zhouchen Lin

Zhitao Hou

Cong Chen

Prashant Ratanchandani

Feng Liang

Chao Bian

Xiangpeng Zhao

Shuo-Hsien (Stephen) Hsiao

Mengyu Zhou

According to implementations of the subject matter described herein, there is provided a solution for automatic recommendation of analysis for a dataset. In this solution, dimension feature information of dimensions of a dataset and operation feature information of candidate analysis operations are extracted. Respective metrics of candidate combinations of the dimensions and candidate analysis operations being suitable for defining an analysis pattern for the dataset are determined based on the dimension and operation feature information. The analysis pattern comprises at least one dimension to be analyzed and at least one analysis operation to be performed on the at least one dimension. A recommendation on an analysis pattern for the dataset is provided based on the determined respective metrics, to indicate a candidate combination. In this way, it is possible to assess and provide a suitable analysis pattern for a given dataset, thereby facilitating quick completion of the data analysis task.

Automatic recommendation of analysis for dataset

Implementations of the present disclosure provide a solution for spreadsheet table transformation. In this solution, one or more header areas and a data area of a spreadsheet table are detected. A hierarchical structure of each of the header areas is determined by analysis of cell merging and/or indents in the header area, and/or a function relationship between data items in corresponding cells of the data area. The spreadsheet table can be transformed to a relational table based on recognition of the hierarchical structure of the header area. In this way, by facilitating understanding of header structures based on the header hierarchy, it is possible to achieve automated transformation from spreadsheet tables to relational tables.

Spreadsheet table transformation

According to implementations of the present disclosure, there is provided a solution for auto-formatting of a data table. A computer-implemented method comprises obtaining values of at least one attribute for a plurality of cells in a data table, the values of the at least one attribute indicating at least one of a semantic meaning of data filled in the cells or a structure of the data table, the cells being arranged in rows and columns in the data table; extracting a feature representation of the values of the at least one attribute In using a first learning network; and determining respective formats for the cells based on the feature representation using at least the first learning network. In this way, it is possible to implement automatic mapping from attribute settings of cells to cell formats by using a learning network, thereby achieving auto-formatting of the data table.

Auto-formatting of a data table

A solution for spreadsheet table transformation is provided. In this solution, one or more header areas and a data area of a spreadsheet table are detected. A hierarchical structure of each of the header areas is determined by analysis of cell merging and/or indents in the header area, and/or a function relationship between data items in corresponding cells of the data area. The spreadsheet table can be transformed to a relational table based on recognition of the hierarchical structure of the header area. In this way, by facilitating understanding of header structures based on the header hierarchy, it is possible to achieve automated transformation from spreadsheet tables to relational tables.

A computing apparatus of an insight interfacing system receives from a user a request for a dataset comprising a plurality of subspaces of a multi-dimensional data structure. Insights are received based on the received request then presented on a display device. Also, a chart may be presented based on the received request. The computing apparatus receives a selection of at least a portion of the presented one or more insights or a portion of the chart, then receives contextual insights based on the selected portion and presents the contextual insights on the display device.

Multi-dimensional data insight interaction

Automatically identifying insights from a dataset and presenting the insights graphically and in natural language text ranked by importance is provided. Different data types and structures in the dataset are automatic recognized and matched with a corresponding specific analysis type. The data is analyzed according to the determined corresponding analysis types, and insights form the analysis are automatically identified. The insights within a given insight type and between insight types are ranked and presented in order of importance. Insights include those having multiple pipelined attributes and other insights include multiple insights identified as having some relationship for the included insights.

Automatic insights for multi-dimensional data

According to implementations of the subject matter described herein, there is proposed a solution for automatic analysis of a difference between multi-dimensional datasets. In this solution, an analysis request is received for a first dataset and a second dataset, each of which including data items corresponding to a plurality of dimensions. In response to the analysis request, data items corresponding to a first dimension in the first and second datasets are compared. Based on the comparison, a first set of influence factors associated with the first dimension are determined, each influence factor indicating a reason for a difference between the first and second datasets from a respective perspective. An analysis result related to the difference between the first and second datasets is presented based on the first set of influence factors. In this way, it is possible to achieve automatic and efficient analysis of the difference between the different datasets.

Automatic analysis of difference between multi-dimensional datasets

The subject matter described herein relates to table detection in a spreadsheet. According to implementations of the subject matter described herein, there is proposed a solution for determining a table in a spreadsheet. In the solution, respective multiple attributes of multiple cells comprised in the spreadsheet may be extracted. Respective features of the multiple cells may be determined based on the extracted multiple attributes. The multiple cells may be divided into at least one candidate area based on the features. At least one candidate table in the spreadsheet may be determined based on the at least one candidate area. By means of the solution, respective features of the multiple cells comprised in the spreadsheet may be determined based on the respective attributes of the multiple cells, and further, a candidate region where a table might exist may be determined based on the respective features of the multiple cells.

Table detection in spreadsheet

Automatically identifying insights from a dataset and presenting the insights graphically and in natural language text ranked by importance is provided. Different data types and structures in the dataset are automatic recognized and matched with a corresponding specific analysis type. The data is analyzed according to the determined corresponding analysis types, and insights from the analysis are automatically identified. The insights within a given insight type and between insight types are ranked and presented in order of importance. Insights include those having multiple pipelined attributes and other insights include multiple insights identified as having some relationship for the included insights.

A system for frequent pattern mining uses two layers of processing: a plurality of computing nodes, and a plurality of processors within each computing node. Within each computing node, the data set against which the frequent pattern mining is to be performed is stored in shared memory, accessible concurrently by each of the processors. The search space is partitioned among the computing nodes, and sub-partitioned among the processors of each computing node. If a processor completes its sub-partition, it requests another sub-partition. The partitioning and sub-partitioning may be performed dynamically, and adjusted in real time.

Frequent pattern mining

Growing community of inventors

Beijing, China

Shi Han