International Business Machines Corporation

Jing-Yun Shyr

Damir Spisic

Yea Jane Chu

Jing James Xu

Sier Han

Daniel Jay Rope

Xueying Zhang

Marc S Altshuller

Fan Li

Weicai Zhong

Richard John Oswald

Margaret J Vais

Graham John Wills

Jean-Francois Puget

Svetlana Levitan

Yea-Jane Chu

Dong Liang

Marius I Danciu

Michael D Woods

Jane Y Chu

Michael McRoberts

Raymond Wright

Neil Whitney

Devendra G Tasgaonkar

Vikremjeet Singh Bhagi

Using a natural language processing (NLP) engine executing in conjunction with a machine that is engaged in first natural language interaction, an analytics intent comprising an analysis type to be performed on a dataset is extracted from the first natural language interaction. Within the dataset, a subset of the dataset comprising data having above a threshold relevance measure with respect to the analytics intent is identified. From the subset, a knowledge graph modeling a set of relationships between data in the subset is constructed. Using the analytics intent and the knowledge graph, a conversational template is customized, augmenting the conversational template with a set of entities corresponding to the analytics intent. To obtain a result, the subset is analyzed using the knowledge graph. A second natural language interaction is presented via the machine, the presenting comprising transforming by the NLP engine the result to fit the customized conversational template.

Natural language interaction based data analytics

A method, system, and/or computer program product analyses event transactional related data to generate insights and predictions, which are pre-created to efficiently respond to requests for prediction/forecasting information, in order to improve the operation of the prediction-generating computer. One or more processors receive a series of structured data, where each entry (E) from the series of structured data has one or more time fields Tand one or more attributes A. In response to determining that the series of structured data is transactional, one or more processors select a time field Tthat meets an aggregation criterion, and then aggregate the transactional data from the time field Tinto a time series data format. One or more processors consolidate results from a time series analysis and a regression analysis of the transformed transactional data to create a consolidated result, which is used to respond to a request for prediction/forecasting information.

Transforming a transactional data set to generate forecasting and prediction insights

A computer-implemented method for prescriptive time-series forecasting, which combines both what-if analysis and goal-seeking analysis. The method comprises building a model for a target metric with a set of predictors, based on historical time-series data, and computing, using the model, a set of forecast values. Using the set of forecast values with respect to a forecasting period, both a set of goals for the target metric and a set of constraints for the predictors are analyzed. A set of updated forecasts based on the analyses with respect to the forecasting period is determined to meet the goals within the set of constraints. The updated set of forecasts is presented with respect to the forecasting period, e.g., using a table, a visualization, and/or an interactive user interface.

System and method for combining what-if and goal seeking analyses for prescriptive time series forecasting

A data portion of a data set utilized in a computerized visualization is analyzed to identify one or more areas of interest each including data values representing distinguishable features relative to the data set. An explanation for the data values of each of the one or more areas of interest is determined. Each explanation is based on other data portions of the data set contributing to the distinguishable features. At least one display layer including labels describing the one or more areas of interest is generated. The labels include the explanation for each of the one or more areas of interest. The at least one display layer is disposed over the computerized visualization to produce an annotated visualization with the labels positioned proximate the one or more areas of interest.

Generation of annotated computerized visualizations with explanations for areas of interest

Techniques are described for automated selection of components for a generalized linear model. In one example, a method includes determining a candidate set of distributions, a candidate set of link functions, and a candidate set of predictor variables, based at least in part on a dataset of interest. The method further includes selecting a distribution from the initial candidate set of distributions and a link function from the initial candidate set of link functions, based at least in part on the candidate set of predictor variables; and selecting predictor variables from the candidate set of predictor variables, based at least in part on the selected distribution and the selected link function. The method further includes reiterating the selecting processes until a stopping criterion is fulfilled, and generating a generalized linear model output comprising the selected distribution, the selected link function, and the selected predictor variables.

Automated selection of generalized linear model components for business intelligence analytics

A method for comparing predictive data models based on a predictive model search is provided. The method may include receiving a first and second portion of a set of data. The method may also include identifying a first and second variation of the second portion, wherein the first variation is different from the second variation. The method may further include generating first predictive data models based on the first variation, and second predictive data models based on the second variation. Additionally, the method may include applying a criteria to rank the first and second predictive data models based on predictive strength. The method may also include presenting a display of the ranked criteria, comprising the first portion, and a portion of the first and second predictive data models, wherein the portion of the first and second predictive data models are collectively ranked and presented according to the predictive strength.

Predictive model search by communicating comparative strength

Techniques are described for generating characterizations of time series data. In one example, a method includes extracting a trend-cycle component, a seasonal component, and an irregular component from a time series of data. The method further includes performing one or more pattern analyzes on the trend-cycle component, the seasonal component, and the irregular component. The method further includes, for each pattern analysis of the one or more pattern analyzes, performing a comparison of an analytic result of the respective pattern analysis to a selected significance threshold for the respective pattern analysis to determine if the analytic result passes the significance threshold for the respective pattern analysis. The method further includes generating an output for each of the analytic results that pass the significance threshold for the respective pattern analysis.

Automatic time series exploration for business intelligence analytics

Techniques are described for generating characterizations of time series data. In one example, a method includes extracting a trend-cycle component, a seasonal component, and an irregular component from a time series of data. The method further includes performing one or more pattern analyses on the trend-cycle component, the seasonal component, and the irregular component. The method further includes, for each pattern analysis of the one or more pattern analyses, performing a comparison of an analytic result of the respective pattern analysis to a selected significance threshold for the respective pattern analysis to determine if the analytic result passes the significance threshold for the respective pattern analysis. The method further includes generating an output for each of the analytic results that pass the significance threshold for the respective pattern analysis.

Provided are techniques for summarizing statistical results. Multiple sets of statistical results are received, wherein each of the sets of statistical results are ordered according to interestingness. Insights are generated based on the multiple sets of statistical results. Relationships between the generated insights are identified. An executive summary is generated with a set of findings based on the identified relationships. An interactive visualization is provided with the generated insights based on the executive summary.

Interpretation of statistical results

One or more processors initiate cluster feature (CF)-tree based hierarchical clustering on leaf entries of CF-trees included in a plurality of subsets. One or more processors, generate respective partial clustering solutions for the subsets. A partial clustering solution includes a set of regular sub-clusters and candidate outlier sub-clusters. One or more processors generate initial regular clusters by performing hierarchical clustering using the regular sub-clusters. For a candidate outlier sub-cluster, one or more processors determine a closest initial regular cluster, and a distance separating the candidate outlier sub-cluster and the closest initial regular cluster. One or more processors determine which candidate outlier sub-clusters are outlier clusters based on which candidate outlier sub-clusters have a computed distance to their respective closest initial regular cluster that is greater than a corresponding distance threshold associated with their respective closest initial regular cluster.

Distributed clustering with outlier detection

A computing device includes at least one processor, and at least one module operable by the at least one processor to receive data representing a hierarchy, wherein the hierarchy comprises at least one set of sibling nodes and a respective parent node, generate a condensed hierarchy by determining a grouping for the at least one set of sibling nodes, determine whether the at least one set of sibling nodes can be represented by the respective parent node, based at least in part on the grouping for the at least one set of sibling nodes, and responsive to determining that the at least one set of sibling nodes can be represented by the respective parent node, remove the at least one set of sibling nodes from the condensed hierarchy. The at least one module may further be operable by the at least one processor to output the condensed hierarchy for display.

Condensing hierarchical data

One or more processors generate subsets of cluster feature (CF)-trees, which represent respective sets of local data as leaf entries. One or more processors collect variables that were used to generate the CF-trees included in the subsets. One or more processors generate respective approximate clustering solutions for the subsets by applying hierarchical agglomerative clustering to the collected variables and leaf entries of the plurality of CF-trees. One or more processors select candidate sets of variables with maximal goodness that are locally optimal for respective subsets based on the approximate clustering solutions. One or more processors select a set of variables, which produce an overall clustering solution, from the candidate sets of variables.

Growing community of inventors

Naperville, IL, United States of America

Jing-Yun Shyr