Interactive user interface for dynamic data analysis exploration and query processing

Abstract

The systems and methods described herein provide highly dynamic and interactive data analysis user interfaces which enable data analysts to quickly and efficiently explore large volume data sources. In particular, a data analysis system, such as described herein, may provide features to enable the data analyst to investigate large volumes of data over many different paths of analysis while maintaining detailed and retraceable steps taken by the data analyst over the course of an investigation, as captured via the data analyst's queries and user interaction with the user interfaces provided by the data analysis system. Data analysis paths may involve exploration of high volume data sets, such as Internet proxy data, which may include trillions of rows of data. The data analyst may pursue a data analysis path that involves, among other things, applying filters, joining to other tables in a database, viewing interactive data visualizations, and so on.

Description

BACKGROUND

The approaches described in this section are approaches that could be pursued, but not necessarily approaches that have been previously conceived or pursued. Therefore, unless otherwise indicated, it should not be assumed that any of the approaches described in this section qualify as prior art merely by virtue of their inclusion in this section.

Exploration, mining, and analysis of high-volume data sources and data warehouses can be a time and resource intensive process, both for the user and for the underlying computing systems which provide data analysis features. Large volumes of data contribute to high consumption of computer memory, and processing of such large volumes of data can require extremely powerful computer processors and processing time. Users such as data analysts often face a seemingly insurmountable task of attempting to locate discrete pieces of information from data sources which can include or primarily consist of innocuous or noisy data which provides little probative value. Often, many iterations to explore such data are required, with each iteration having an associated cost in time and resources.

BRIEF DESCRIPTION OF THE DRAWINGS

FIGS. 1A-1M illustrate an example data analysis workflow performed using several interactive data analysis user interfaces, as generated using one embodiment of the data analysis system of FIG. 6.

FIGS. 2A-2B illustrate example data analysis path tree view user interfaces, as generated using one embodiment of the data analysis system of FIG. 6.

FIG. 3 is a flowchart illustrating one embodiment of an example process for processing data set queries using an interactive data analysis user interface, as used in one embodiment of the data analysis system of FIG. 6.

FIG. 4 is a flowchart illustrating one embodiment of an example process for generating an interactive breadcrumb and/or an interactive tree view representing a data analysis path for display in a data analysis user interface, as used in one embodiment of the data analysis system of FIG. 6.

FIG. 5 is a flowchart illustrating one embodiment of an example process for navigating a data analysis path using an interactive breadcrumb and/or an interactive tree view representing the data analysis path in data analysis user interface, as used in one embodiment of the data analysis system of FIG. 6.

FIG. 6 is a block diagram of an implementation of an illustrative data analysis system.

DETAILED DESCRIPTION

Overview

Data analysts face a daunting challenge when tasked with exploring a large volume of data to identify discrete pieces of information and intelligence—a veritable needle-in-a-haystack problem. In some cases a data analyst may have a lead on something she wishes to investigate. For example, knowing that a particular bad actor was involved in an incident may be a starting point from which the data analyst can begin to investigate other individuals who may have been involved. However, in many cases the data analyst has little information or, at best, only some contextual information on which to base an investigation. For example, the data analyst might know that a website was hacked at a certain time but have little else to go on to identify the individuals involved. However, data analysts might need to explore multiple analysis paths, many of which lead to “dead ends” or results that do not answer the search premise, by sorting and filtering through a massive volume of (for example) Internet proxy traffic data, cyber security data, telephone data, and/or any other type of data. The data analyst's path through the data is often multi-directional, in that the analyst may approach a problem from the top-down, bottom-up, and/or in either direction over the course of a given data analysis path and investigation. For example, the data analyst may begin with a large set of data, submit one or more queries which may lead to smaller sets of data results or filtered data results, and then merge or join smaller sets of data results to formulate larger sets of data, exclude subsets of data, and so on in as many iterations as the data analysts desires or finds useful.

The systems and methods described herein provide highly dynamic and interactive data analysis user interfaces which enable the data analyst to quickly and efficiently explore large volume data sources. In particular, a data analysis system, such as described herein, may provide features to enable the data analyst to investigate large volumes of data over many different paths of analysis while maintaining detailed and retraceable steps taken by the data analyst over the course of an investigation, as captured via the data analyst's queries and user interaction with the user interfaces provided by the data analysis system. For example, one data analysis path may begin with a high volume data set, such as Internet proxy data, which may include trillions of rows of data corresponding to Internet communication events even over a short period of time. The data analyst may pursue a data analysis path that involves applying a filter, joining to another table in a database, viewing a histogram, manipulating a pivot table, and finally viewing a summary of the entries located which match the parameters of the data analysis path. The summary may not provide the data analyst with the results she was hoping to find and thus she may decide to investigate another data analysis path. However, rather than start over from scratch, she may wish to return to another point in the first data analysis path. For example, perhaps while viewing the histogram she noted other points to explore later.

In an embodiment, a main data analysis workflow user interface provided by the data analysis system provides the data analyst with a set of highly interactive data visualization tools with which the data analyst can explore sets of data. Data visualization tools provided by the data analysis system can include, for example, a data summary view; a time series view; a time grid view; a heat map view; a pivot table view; an ability to link to another table in a database; and options to add, remove, or keep only selected results from a data view.

Each of these data visualization tools may include rich UI controls which are configured to respond to user interaction and input in a manner that allows the non-technical user to quickly and intuitively create complex data set queries. For example, in a histogram view data visualization UI, when a user selects (e.g., by clicking on, touching, speaking, etc.) a column in the histogram, a query filter may be automatically determined and applied to a subsequent query to be processed by the data analysis system. Or, in another example, an interactive heat map overlaid on a geographic region may allow the user to click and/or touch-and-drag to select a region of interest, and in response a query filter may be automatically generated which specifies the region of interest as a matching parameter. Subsequent queries may be easily built and executed to automatically apply the filter specified by the user interaction with the various data visualizations, providing a highly responsive, intuitive, and immersive user experience.

In one embodiment, the data analysis system maintains a session history for the analyst's data analysis paths and may provide various user interface features to allow the analyst to quickly and easily return to any prior point in the data analysis path, including the histogram view in the hypothetical example above. When a new data analysis path is started the session history is updated to include a new branch path from the original path.

As the analyst builds a data set query using these data visualization tools, the corresponding query result is added to the main data analysis workflow UI (for example, appended to the bottom of the workflow), and the session history is updated. This analysis interaction process may proceed over as many data set queries as the analyst wishes to investigate, with each subsequent query result added to the main data analysis workflow UI and the session history being updated.

In one embodiment, the main data analysis workflow UI provides the ability for the data analyst to scroll up or down within the current data analysis path, for example to re-display various data set query results and associated data visualizations appearing earlier in the current data analysis path (which, for example, may no longer fit in a displayable area of a display screen). The data analyst user may then decide to investigate a new data analysis path based on an earlier query result/visualization.

To assist the data analyst user in navigating these multiple data analysis paths, data query sets, and results, the data analysis system can provide a variety of path navigation user interface tools. For example, in one embodiment, an interactive breadcrumb path UI control may be generated and displayed (e.g., at the top of the main data analysis workflow UI, as shown in the example user interfaces herein; or in any position). Each “crumb” in the breadcrumb path UI control may indicate a summary of a respective data query and/or filter submitted over the course of the current data analysis path that is displayed in the main data analysis workflow UI. Further, each crumb may be selectable by the data analyst. Selecting a crumb may automatically update the main data analysis workflow UI to re-display the associated data set query and data visualization of the associated results.

In another embodiment, an interactive tree view UI control may be generated and displayed (e.g., at the right side of the main data analysis workflow UI, as shown in the example user interfaces herein; or in any position). Each “node” in the tree view path UI control may indicate a summary of the respective data queries submitted over the course of the current data analysis path that is displayed in the main data analysis workflow UI. For example, when a user positions a cursor or pointer over a node in the tree view, a popover UI element may appear next to the node, where the popover UI element includes a summary of the data query. Further, each node may be selectable by the data analyst. Selecting a node may automatically update the main data analysis workflow UI to re-display the associated data set query and data visualization of the associated results. The data analyst may then, for example, jump to another point, save the associated data set, or begin a new branched data analysis path.

The tree view UI control may be displayed in a variety of ways. In one embodiment, a single-path tree view UI control may be displayed corresponding to the current data analysis path. For example, suppose the current data analysis path (and the main data analysis workflow UI) includes a filter with a histogram view; a join with a table view; and a filter with a time series view. Then, the single-path tree view UI control may include one path comprising three nodes: filter-join-filter. In some embodiments the different associated data visualization elements may be indicated by representative icons, such the data analyst can view at-a-glance that the current analysis path comprises filter-join-filter.

In another embodiment, an all-paths tree view UI control may be displayed to present the user with all analysis paths explored for the current session. For example, suppose the current data analysis path is as described above with two additional branch paths: a heat map view branched from the join; and a time grid view branched from the first filter. Then, the all-paths tree view UI control may include five nodes corresponding to the three paths: filter-join-filter (the current path); filter-histogram; and filter-time grid. In one embodiment, the single-path tree view UI control may be displayed concurrently with the main data analysis workflow UI and display the current data analysis path. A user-selectable option to view the all-paths tree view UI control may be presented, and upon selection the all-paths tree view UI control may displayed such as in a popover UI element. The all-paths tree view UI control may be configured to support full user interaction in a similar manner to the single-path UI control. Thus, the data analyst may switch to the all-paths tree view UI control, select a node from the all-paths tree, and in response the main data analysis workflow UI will re-display the associated data set query and data visualization of the associated results.

The breadcrumb path UI control and the tree view path UI control can thus each be used to jump from any first point in the data analysis path to any second point in the data analysis path, with the main data analysis workflow UI being updated to match the selected point. However, in certain embodiments, when the user jumps from point to point in this manner, the associated data query may not need to be executed again. Rather, a cached copy of the previously generated data visualization may be re-displayed, or a cached copy of the associated data query results may be accessed and used to generate or update the data visualization.

The data analysis system described herein provides a variety of additional features and benefits. In certain embodiments, the data analysis system and provided user interfaces may enable the data analyst user to save data query set results at any stage in the analysis path; and/or export data query set results, for example to another database or another format such as a presentation software format (e.g., PowerPoint), as further described below.

Saving a data query set may allow the data analyst to return to an analysis path in progress at a later time for further exploration. In some instances the data query set may be saved such that the full analysis path taken to reach the particular data query set is saved as well, which can enable the analyst to re-explore the full path at a later time. In some cases, the data query set may be saved in a “clean” state, such that the analysis path taken to reach the particular data query set is not saved. This may be of benefit, for example, if the analysis path was particularly long or convoluted and saving the path provides little value to the analyst. The system may provide an option for the analyst to choose whether a data set should be saved with or without the full analysis path.

Several types or varieties of data sets may be saved in accordance with this feature. One example data set is a materialized saved set, which is a particular set of records that matches the exact set of records that an analyst currently observes, and can be retrievable in the same precise form at a later date. A materialized saved set may have different versions that are refreshable or updatable at a later date, and as such information can also be stored describing, among other things, how a refresh may be performed, when the last refresh was conducted, and who performed the refresh. Another example data set is a dynamic saved set, which is a workflow definition that will produce a set applying some saved criteria. Another example data set is a parameterized saved sets, which is a workflow definition that will produce a set applying some saved criteria against with sets of the same shape (e.g., column definitions) as the original saved definition input sets. Many other types of saved sets may be possible as well.

Exporting a data query set may allow the data analyst to extract a data query set, for example, for importing into another data source separate from the original, high-volume data source. This feature may be of particular benefit if, for example, the original, high-volume data source comprises data which is very noisy or may otherwise be of little value in more focused or specialized databases. Using the example above with reference to Internet proxy traffic data, the vast majority of such data is normal, everyday traffic which is likely not to be of particular interest. Rather than import the full set of Internet proxy data into a more specialized database, the data analyst can use the data analysis system herein to explore and filter down the high-volume data. Once a filtered data query set is obtained which appears to be of interest or relevance to the specialized database, the data set can be exported.

Another benefit to the export feature is that it may reduce the costs involved in maintaining and searching a high-volume data source by keeping that data source separate. For example, the cost (both financially and physical resource-wise in terms of memory, response time, processing speed, efficiency, power consumption, etc.) of maintenance, duplication, backups, and other redundancy measures for specialized databases can grow exponentially if copious amounts of noisy data are imported. Thus, using the data analysis system and export features herein, the specialized databases can remain specialized, while the high-volume data sources may still be explored to identify and extract valuable data.

The data analysis systems and methods described herein may also include various features for query optimization on the backend or server-side. For example, a query backend module may be configured to analyze user selections and interactions with the data set query results and associated data visualizations and, based on those selections and interactions, determine what data the user is actually interested in, and translate the query into an efficient or optimized search query string. For example, the user might produce a workflow or analysis path which involves linking or joining a first large data set to a second large data set, and then filtering the resulting linked data set. An optimized search query string may then be generated which applies the filter operation prior to the link or join operation, such that the resources required to perform the join operation would be greatly reduced (i.e., since the link or join operation would be performed on smaller data sets resulting from the filter being applied first).

Another query optimization feature which may be provided by the data analysis system includes logic to evaluate a query, such as a JOIN operation, will be and, based on the expense, determine whether a temp table should be created in order to improve the performance of subsequent queries using the data query result set. For example, in some cases a temp table may “cost” less in terms of computer processing or memory requirements than re-executing the JOIN operation. Another query optimization feature related to the use of such temp tables which may be provided by the data analysis system includes connection pooling, whereby temp tables may be mapped to different connections or users to facilitate efficient re-use of temp tables for the respective user's analysis workflow. These and other features described herein are of particular value in the context of processing and querying high-volume data, where physical hardware costs and resources are constrained.

In some embodiments, the data analysis system may be configured to defer execution of a query until a particular analytic view is rendered. This can provide additional efficiency, for example, when users are simply trying to construct complex set-based queries, and enables non-technical users to create complex queries without actually understanding the underlying query language (like SQL). Users might find benefit from the data analysis system as a tool that enables construction of complex reporting queries, even without visual feedback (but with the benefit of optionally including such visual feedback on demand). In some instances, the data analysis system may provide users with the ability to define filters without visual feedback, and do so incrementally and intuitively as a logical “next step” in a workflow for filtering, triaging, linking, and so on.

Example User Interfaces

FIGS. 1A-1D and 2A-2B illustrate example user interfaces and UI components, as used in one or more embodiments of the data analysis system 100 of FIG. 6. The sample user interfaces may be displayed, for example, via a web browser (e.g., as a web page), a mobile application, or a standalone application. In some embodiments, the sample user interfaces shown in FIGS. 1A-1D and 2A-2B may also be displayed on any suitable computer device, such as a cell/smart phone, tablet, wearable computing device, portable/mobile computing device, desktop, laptop, or personal computer, and are not limited to the samples as described herein. The user interfaces include examples of only certain features that a data analysis system may provide. In other embodiments, additional features may be provided, and they may be provided using various different user interfaces and software code. Depending on the embodiment, the user interfaces and functionality described with reference to FIGS. 1A-1D and 2A-2B may be provided by software executing on the individual's computing device, by a data analysis system located remotely that is in communication with the computing device via one or more networks, and/or some combination of software executing on the computing device and the data analysis system. In other embodiments, analogous interfaces may be presented using audio or other forms of communication. In an embodiment, the interfaces shown in FIGS. 1A-1D and 2A-2B are configured to be interactive and respond to various user interactions. Such user interactions may include clicks with a mouse, typing with a keyboard, touches and/or gestures on a touch screen, voice commands, physical gestures made within a proximity of a user interface, and/or the like.

FIG. 1A illustrates an example data analysis user interface 1000A, as generated using one embodiment of the data analysis system 100 of FIG. 6. The data analysis UI 1000 may comprise, for example, a main data analysis workflow user interface for a user, such as a data analyst, to dynamically build data sets and queries and to view query set results. The query set results may be generated by the data analysis system 100 in response to user queries, and presented via highly interactive user interface UI display elements and controls. The data analysis UI 1000 may include a tree view UI control 102A, breadcrumb UI control 104A, and a data visualization UI element 106, which in FIG. 1 comprises a summary data view presenting a summary of data set results. Also shown is an action menu 108 which provides a list of data analysis tools and options available for the user, including: summary view; histogram; time series; time grid; heat map; pivot table; a set viewer; a table view; a pivot table; and options to save a set and to export a set. Additional actions and options not shown can also be included. In response to selection of these various data analysis tools the data analysis UI 1000 may be updated to add a new UI display panel presenting the selected analysis tool at the bottom of the workflow area. The user may then use the selected analysis tool to build and run a subsequent query to obtain a next data query result set.

Each time a selected data analysis tool is used by the user to build a query, the data analysis system 100 may update the associated session history for the user to track the user's analysis path. In some embodiments, the session history may be updated as the user builds the query, but before the query is executed, in order to capture the user's work in progress. In some embodiments, the session history may be updated when the query is executed and query results are obtained. As the session history is updated, the tree view UI control 102A and the breadcrumb UI control 104A may also be updated to reflect the updated analysis path. This is described in more detail with reference to FIGS. 1B-1D as the example workflow progresses; as well as with respect to the process 400 of FIG. 4 herein. For example, the breadcrumb UI control 104A may be appended with a description of the query parameters and/or type of operation performed (e.g., as may be indicated by an icon). Similarly, the tree view UI control 102A may be updated to include an additional node on the current branch or path, corresponding to the additional analysis step (including either a query-in-progress or a query result set generated and added to the main workflow).

As the analysis path grows or the user begins to backtrack and create new branches, the tree view may become too large to persistently display in the main data analysis workflow user interface. Thus, in some embodiments, the tree view UI control 102A-D may comprise a single series of nodes representing the current analysis path in the session. The tree view UI control 102A-D may include an expand option which the user can select to view the full or expanded tree view. The full or expanded tree view may be presented as a popover UI control overlaid on the main data analysis workflow user interface or in a separate window or user interface. More detailed examples of the expanded tree view are illustrated and described with respect to FIGS. 2A and 2B herein.

With continued reference to FIG. 1A, the hand cursor over the histogram icon in the actions menu 108 indicates that the user is selecting a histogram data visualization, which is illustrated and described next in FIG. 1B.

FIG. 1B illustrates an example data analysis user interface 1000B, which may be a continuation of the workflow shown in user interface 1000A. Once the user has selected the histogram option, a histogram menu 110 may be displayed by which the user can select one or more query parameters specifying the desired output result and format. Other data visualization elements described herein may have similar or other menu options for the user to specify query parameters for the respective data visualization.

When the user is satisfied with the selection, e.g., of count_distinct, hometown, country, and quantity parameters in the example of FIG. 1B, the “compute” button may be selected. In response the data analysis system 100 receives the request and query parameters and processes a query against the current query result set. The output results may be used to generate the histogram data visualization and displayed in the main data analysis workflow user interface, as FIG. 1B shows. Also shown in user interface 1000B, the view UI control 102B may be updated to add an additional node indicating the histogram has been added to the current analysis path. In some embodiments, the visualization, e.g., the histogram, is updated in real time as the various parameters are each individually selected.

The histogram data visualization may be interactive and provide functionality for the analyst to quickly and intuitively begin developing another query. For example, in response to the user selecting one or more bars in the histogram, a query filter display 114B may appear and dynamically update to include each of the one or more bars selected by the user. When the user wishes to continue the analysis beyond the histogram, another action may be selected from the action menu 108B. If any query filter criteria are still selected from the user interaction with the histogram, these may be gathered and applied to the next query processed by the data analysis system for the current workflow.

FIG. 1C illustrates an example data analysis user interface 1000C, which may be a continuation of the workflow shown in user interface 1000B. In this continued example of the workflow which started in FIG. 1A, the data analyst has progressed a couple of steps further in the analysis path. The view UI control 102C has been updated to indicate that subsequent to the histogram discussed above, a time series query was processed by the system and the associated time series data visualization (shown partially visible in the user interface 1000C) was generated and displayed. At any point in the analysis the user may use the fully interactive view UI control 102C to navigate to an earlier step in the analysis path; in response, the main data flow user interface may scroll up or otherwise cause re-presentation of the data visualization corresponding to the selected node.

The user interface 1000C also demonstrates the breadcrumb UI control 104C being updated to reflect the user's earlier selection from the histogram to filter on the country “USA.” As with the tree view UI control 102C, the user may select any crumb in the breadcrumb UI control 104C to navigate to the corresponding step in the analysis path.

The time series data visualization, like the histogram, may provide several interactive features. For example, in response to the user selecting a segment of the time series, the query filter display 114C may dynamically update to specify a query filter corresponding to the user selection (here, a date range). As user interface 1000C further illustrates, a summary view has been selected again by the data analyst so that she can assess at a glance whether the query filters applied so far in the analysis path have yielded any interesting query results.

FIG. 1D illustrates an example data analysis user interface 1000D, which may be a continuation of the workflow shown in user interface 1000C. In this continued example of the workflow which started in FIG. 1A, the analyst has selected a heat map data visualization element 116. The view UI control 102D has been updated to indicate that subsequent to the histogram and time series discussed above, a summary view, a pivot table, and finally the heat map currently displayed were processed by the data analysis system. As before, at any point in the analysis the user may use the fully interactive view UI control 102D to navigate to an earlier step in the analysis path.

The user interface 1000D also demonstrates the breadcrumb UI control 104C being updated to reflect the user's earlier selection from the histogram to filter on the country “USA,” as well as several additional interim filters and operations which have been applied during the current analysis path. As with the breadcrumb UI control 104C, the user may select any crumb in the breadcrumb UI control 104D to navigate to the corresponding step in the analysis path.

The heat map data visualization, like the histogram and the time series, may provide several interactive features. For example, in response to the user drawing a circle or otherwise circumscribing an area on the map, the query filter display 114C may dynamically update to specify a query filter corresponding to the user selection (here, a geographic range and a center point).

In the example shown, the user has selected a region roughly corresponding to the New England area, with New York as the center focus. For instance, the query results data may indicate a high degree of activity, such as Internet traffic, satisfying the query parameters in the New York area. The data analyst/user may quickly make this determination from studying the heat map, and then uses the map selector to zoom in on the region of interest. In response the data analysis user interface may interpret the user's map selection as a desire to build a new query and immediately add a query filter, shown below the map in this example, to filter the results to the selected region. The user can then select one of the actions 108D and execute the query filter to view a subsequent data set, where the filter is applied to the query results from the heat map.

FIGS. 1E-1M illustrate additional user interface features that may be implemented in some embodiments in order to selection actions, visualizations, and/or data sets upon which new visualizations should be generated. In this example, a text input control 123 receives text entry from the user in a natural language format; although in other embodiments the text entry may be in a coded, shorthand, or abbreviated format. The text entered by user may then be parsed by the system in order to predict another visualization of interest to the user and/or further filter parameters of interest to the user, such as that might be applied to another visualization.

For example, in the example of FIG. 1E, the user has begun typing a query such as “show me count by day.” However, before finishing the text, each of the visualizations shown below the input control 123 are grayed out (e.g., not selectable) as the system is processing the text data to determine which visualization is associated with the provided text. In some embodiments, the system provides suggestions of visualizations/or filtering associated with those visualizations before an entire search string is entered. For example, the system may predict characters that are going to be typed by the user. Returning to the example in the figures, when the text has been entered, as in FIG. 1F (or prior to completion of the full text query in some implementations), the dropdown list of visualizations is updated to provide one or more predicted or suggested data visualization elements relevant to the provided text entry. In the example of FIG. 1F, The prediction or suggestion may be based on a real-time analysis of the user input and scanning for certain keywords which may indicate or otherwise suggest that the user is interested in a particular set of data which may be well suited for one or more data visualization types. For example, the word “count” in the user input may signal to the system that the user is probably interested in viewing a time series, and thus this particular data visualization option or thumbnail in the dropdown list may be highlighted so that the user can see that the “time series” data visualization is recommended and/or available for the particular user input or query. As another example, the user input may be “histogram rate code” which may in turn signal the system that the user is probably interested in viewing a histogram; then the histogram option or thumbnail may be highlighted for the user to more easily select.

The user can then select any of the available visualizations (e.g., time series in example of FIG. 1) for display in the user interface, such as is shown in FIG. 1G. FIG. 1G also includes an input control 123B that provides the same or similar functionality to the input control 123; that is, free-form text may be entered in order for the system to predict which visualization the user is interested in viewing next and/or additional parameters for filtering the data set in the recommended visualizations. However, the visualization prediction provided in response to text provided into input control 123B may further be influenced by the time series that was previously displayed in response to the previous selection made with reference to FIG. 1G, for example. Thus, as with other examples discussed herein, filtering of the data set and subsequent visualizations based on such filtering may build upon one another in a sequential fashion to provide various granularities, types, and sets of information to the user in various available visualizations.

In the example of FIG. 1H, the user has started typing a new free-form search query, but the system has not yet identified a most appropriate visualization and/or parameters for the visualization. Moving to FIG. 1I, the user has completed entering “histogram rate code”, such that a histogram visualization is unlocked for selection, and the system has automatically determined that the “rate code” should be the main input parameter of the histogram. Thus, when the Histogram visualization is selected, a new visualization, such as in FIG. 1J, including a histogram showing rate counts, is added to the user interface.

FIGS. 1K-1M illustrate yet another example of parsing of free-form text in order to provide a particular visualization and primary attribute for display in the visualization. In this example, the user has indicated that a distribution visualization is desired, such that the system automatically begins precalculating (in some embodiments) a distribution visualization based on a “trip distance” parameter of the data set, resulting in display of the distribution visualization of FIG. 1M in response to the user's selection of the Distribution visualization (in FIG. 1L).

In some instances, the dropdown list (or other user interface display element listing one or more options for user selection) may display or include a hint for the user to indicate one or more keywords or data types which, if included in the text input, would enable or highlight the particular data visualization element. For example, the list may indicate that, for a pivot table, the system is “looking for” the keyword “across.” Then, if the user inputs the term “across” in the text input control, the system may infer that the user is interested in viewing a pivot table and correspondingly enable or recommend the pivot table data visualization. Other examples may include: Grid is looking for the keyword “an aggregate”; Histogram is looking for the keyword “hist”; Distribution is looking for the a data type of “a column,” and Time Series is looking for the keyword “time.”

Additional keywords may be used or suggested, and more than one keyword may be “looked for” by a particular data visualization to signal the system to generate, enable, or otherwise recommend the particular data visualization to the user. In addition, the data visualization highlighted or recommended by the system may change in real-time as the user provides more input. For example, the system may initially recommend one data visualization such as a Time Series in response to identifying the keyword “time” in the user input; but then update to recommended a different data visualization such as Pivot Table in response to identifying other keywords or structure to the user input that may indicate the user is more likely interested in something other than the initially recommended Time Series data visualization.

FIGS. 2A-2B illustrates example data analysis path tree view user interfaces, as generated using one embodiment of the data analysis system 100 of FIG. 6. The tree view 202A of FIG. 2A or the tree view 202B of FIG. 2B may be displayed in response to a user selection from the main data workflow user interface. Nodes in each tree view 202A, 202B may be represented by icons indicating the types of query operations or data visualization performed or accessed at the particular node. For example, the tree view 202A includes a summary node, a histogram node, a time series node, a pivot table node, and a summary node on each of two branches from the pivot table node. The tree view 202A may correspond, for example, to some of the analysis path and tree view 102C described in FIG. 1C.

In FIG. 2B, the tree view 202B corresponds to the same analysis path represented in tree view 202A but includes several additional nodes, indicating that the analysis path has grown. For example, the tree view 202B includes the same nodes as tree view 202A, with additional nodes for each of the two branches—namely a table view node and a histogram node.

As further illustrated in FIG. 2B, popover UI elements 204B may be displayed to show query detail (e.g., the associated query parameters and/or type of operation performed). The popover UI elements 204B may be displayed to the right (or left, etc.) of the tree, such as when the user positions a pointer or cursor over a node in the tree view 202A. In some embodiments, only some nodes of tree view 202B display the associated query detail, perhaps indicating certain key nodes in the analysis path. For example, a node's query detail might be selected for display based on a user flag or marker indicating the query result was one of interest for later data exploration; or, the query detail might be selected for display based on a determination made by the data analysis system, such as a based on a number of results returned, or other factors.

In certain embodiments, the tree views 202A, 202B may be interactive such that the user may select a node in order to return to that point in the analysis path. Selection of the node may cause the main data analysis workflow user interface to re-display the associated data query set result and/or visualization control, which the user can view and optionally from which the user can launch a new analysis path.

Although the tree views 202A and 202B shown in FIG. 2A-2B illustrate two branched analysis paths for the current session, in certain embodiments other tree views may be generated to show any number of paths. For example, in one embodiment, a single-path tree view UI control may be presented to show only the current analysis path, for example in a side panel on the main data analysis workflow user interface. An expanded or all-paths tree view UI control may then be displayed in response to user selection of an option presented in one of the data system user interfaces to view the expanded or all-paths tree. In other instances, depending on the depth and complexity of the analysis path, the tree view UI may be generated to display only a local portion of the analysis path, or include a zoom in/zoom out feature to allow the user to navigate the tree view and different analysis path.

In various embodiments, the tree view represents a computation graph (for example, Directed Acyclic Graph) of input sets. Some operations take as input more than just the parent set, and thus form the nodes of a compute graph. Other compute graphs can be used as input at these junctions, and with the added feature of constructing saved sets, these compute graphs can be quite complicated. Thus, one benefit provided by the tree views discussed herein is to allow non-technical users to construct large compute graphs with semantically labeled sets. The data analysis system may be configured to perform some caching on the back-end to optimize the compute profile of particular graphs. This caching can be performed across multiple users, such that if multiple users request the same graph, the result of that compute graph may be cached for speedier or more efficient execution. Elements of the total computation may be optimized by replacing a particular node in the graph with its materialized saved set result (for example, in cases in which the materialization data set remains valid).

Examples of Processes Performed by Data Analysis Systems

FIGS. 3, 4, and 5 are flowcharts illustrating various embodiments of data analysis system processes. In some implementations, the processes are performed by embodiments of the data analysis system 100 described with reference to FIG. 6 and/or by one of its components, such as the network service module 124, the caching module 126, the export service 128, the import service 132, and/or the query backend 134 shown in FIG. 6. For ease of explanation, the following describes the services as performed by the data analysis system 100. The example scenarios are intended to illustrate, but not to limit, various aspects of the data analysis system 100. In one embodiment, the processes can be dynamic, with some procedures omitted and others added.

FIG. 3 is a flowchart illustrating one embodiment of a process 300 for processing data set queries using an interactive data analysis user interface, as used in one embodiment of the data analysis system 100 of FIG. 6. Depending on the embodiment, the method of FIG. 3 may include fewer or additional blocks and/or the blocks may be performed in an order that is different than illustrated.

At block 305, the data analysis system 100 processes a data set query on a data set. The data set may be an unfiltered data set, such as a set of proxy data logs that the user wishes to analysis or may be a filter portion of a data set, such as a subset of the proxy data that was previously processed.

At block 310, the data analysis system 100 generates and provides output results for the data set query, which may include an associated data visualization UI element such as those described herein. The output results and data visualization UI element may be displayed, for example, via one of the example user interfaces 1000A-D discussed above.

At block 315, the data analysis system 100 receives user interaction data for user interaction with the data visualization UI element. For example, the user may interact with the data visualization UI element by selecting one or more bars, lines, or other portions indicating a selection of some subset of the data represented by the data visualization. The selections may then be interpreted as query filters for a subsequent query result set. Several examples of these types of interactions are described with reference to FIGS. 1A-1D.

At block 320, the data analysis system 100 initializes generation of the subsequent query based on the user interaction data. For example, a query filter display may be presented and dynamically updated to include each of the user's selections. Once the user's selections have been received and the user elects to calculate or submit the query, the process 300 may return to block 305 and repeat again. In this way, the user can drill down (and/or up) in various manners to view different subsets of the data in various ways.

FIG. 4 is a flowchart illustrating one embodiment of a process 400 for generating an interactive breadcrumb and/or an interactive tree view representing a data analysis path for display in a data analysis user interface, as used in one embodiment of the data analysis system 100 of FIG. 6. Depending on the embodiment, the method of FIG. 4 may include fewer or additional blocks and the blocks may be performed in an order that is different than illustrated.

At block 405, the data analysis system 100 receives a request to execute a data set query on a current session or result set. The request may be received, for example, via one of the user interfaces 1000A-D described previously, for example over the course of an analysis workflow.

At block 410, the data analysis system 100 executes the data set query, for example by using the caching module 1226 and/or accessing the local data source 170 to run the query. Some or all of the query execution may be performed by, for example, the query backend 134, which may apply some of all of the query optimization techniques described herein.

At block 415, the data analysis system 100 accesses the output results and generates a UI element (e.g., a data visualization control) associated with the execution of the data set query. For example, the type of UI element generated may be based on a user selection from the actions menu 108 illustrated and described with reference to FIG. 1A.

At block 420, the data analysis system 100 maintains or updates a session history of the data set queries executed for the current session. The session history includes the query details associated with each respective data set query, such as query parameters and/or a type of operation performed. The query details may be used, for example, in order to provide a visual reminder to the user about the data set query, as well as to enable re-execution of the query in some instances.

At block 425, the data analysis system 100 generates or updates an interactive breadcrumb UI control (such as the breadcrumb 104A-D to display the summary of the query detail information for data set queries performed for the current analysis path.

At block 430, the data analysis system 100 generates or updates an interactive tree view UI control (such as the tree view 102A-D or 202A-B), which may be configured to optionally display the summary of the query detail information for data set queries performed for the current analysis path (e.g., the query detail information may be displayed in response to a user interaction such as positioning a cursor, gesturing with a finger or hand near the screen, etc.).

Once the interactive breadcrumb and/or tree view UI controls have been updated, the user interfaces 1000, 2000 may be updated or refreshed to display the current view, and the user may continue to build and execute data set queries.

FIG. 5 is a flowchart illustrating one embodiment of a process 500 for navigating a data analysis path using an interactive breadcrumb and/or an interactive tree view representing the data analysis path in data analysis user interface, as used in one embodiment of the data analysis system 100 of FIG. 6. Depending on the embodiment, the method of FIG. 5 may include fewer or additional blocks and the blocks may be performed in an order that is different than illustrated.

At block 505, data analysis system 100 provides for display in a data analysis workflow user interface, an interactive breadcrumb and/or tree view UI element each of which may display summary information about data set queries performed for the current session, analysis path, or result set.

At block 510, data analysis system 100 receivers a user selection of a session history node, such as a crumb in the breadcrumb or a node in the tree view.

At block 515, data analysis system 100 updates the main workflow user interface to display the data set query and results corresponding to the selected session history node.

At block 520, data analysis system 100 updates the interactive breadcrumb to display the data set queries or detail associated with the selected session history node.

At block 525, data analysis system 100 updates the interactive tree view to provide an indication that the current position in the analysis path has changed, and to optionally display a new branch for the current session.

At block 530, data analysis system 100 provides or updates the main workflow user interface as generated at block 515, the interactive breadcrumb as updated at block 520, and the interactive tree view at block 525.

Example System Implementation and Architecture

FIG. 6 is a block diagram of one embodiment of a data analysis system 100 in communication with a network 160 and various systems, such as client computing systems(s) 168, external data source(s) 172. The data analysis system 100 may be used to implement systems and methods described herein, including, but not limited to the process 300 of FIG. 3, the process 400 of FIG. 4 and the process 500 of FIG. 5.

Data Analysis System

In the embodiment of FIG. 6, the data analysis system 100 includes a network service module 124, a caching module 126, an export service 128, an import service 130, a query backend 134, and a local data source 170.

The security/single sign-on manager 122 provides capabilities to ensure that users are properly authenticated with the data analysis system 100, as well as other security measures and protocols which may be used by the data analysis system 100.

The network service module 124 provides capabilities related to generation, presentation, and in some cases display of the user interfaces, UI components, and related features, such as those illustrated and described with reference to FIGS. 1A-1D and 2A-2D herein.

The caching module 126 provides capabilities related to, among other things, caching of query results generated for analysis workflow sessions. The caching module 126 may also be configured to manage connection pooling for the data analysis system 100, such that users of the system are mapped to temp tables created for their respective analysis workflows. Among other benefits this may help improve response times for the users as they may utilize the same connection and same temp tables many times over a single session.

The export service 128 processes requests to export data from the data analysis system 100 to, for example, the external data sources(s) 172, working in association with the caching module 126 to access data and provide it to the network service module 124 for export.

The import service 132 may periodically poll or be notified by the external data source(s) 172 when new data may be available for import. When new data is available for import, the import service 132 may update the local data store 170 and provide an indication to the caching module 126 that new data is available. In response the caching module 126 may purge or invalidate outdated data.

The query backend 134 provides capabilities related to query optimization including analyzing query filters received from the network service module 124.

Each of these components may be stored in a mass storage device as executable software codes that are executed by a central processing units (“CPU”). These and other modules in the data analysis system 100 may include, by way of example, components, such as software components, object-oriented software components, class components and task components, processes, functions, attributes, procedures, subroutines, segments of program code, drivers, firmware, microcode, circuitry, data, databases, data structures, tables, arrays, and variables. In the embodiment shown in FIG. 6, the data analysis system 100 is configured to execute the modules recited above to perform the various methods and/or processes herein (such as the processes described with respect to FIGS. 3, 4, and 5 herein).

The data analysis system 100 and/or its subcomponents may include, for example, a server, workstation, or other computing device. In one embodiment, the exemplary data analysis system 100 includes CPUs, which may each include a conventional or proprietary microprocessor. The data analysis system 100 further includes one or more memories, such as random access memory (“RAM”) for temporary storage of information, one or more read only memories (“ROM”) for permanent storage of information, and one or more mass storage devices, such as a hard drive, diskette, solid state drive, or optical media storage device. Typically, the modules of the data analysis system 100 are connected to the computer using a standard based bus system. In different embodiments, the standard based bus system could be implemented in Peripheral Component Interconnect (“PCI”), Microchannel, Small Computer System Interface (“SCSI”), Industrial Standard Architecture (“ISA”), and Extended ISA (“EISA”) architectures, for example. In addition, the functionality provided for in the components and modules of data analysis system 100 may be combined into fewer components and modules or further separated into additional components and modules.

The data analysis system 100 and/or its subcomponents are generally controlled and coordinated by operating system software, such as Windows XP, Windows Vista, Windows 7, Windows 8, Windows Server, UNIX, Linux, SunOS, Solaris, iOS, Blackberry OS, or other compatible operating systems. In Macintosh systems, the operating system may be any available operating system, such as MAC OS X. In other embodiments, the data analysis system 100 may be controlled by a proprietary operating system. Conventional operating systems control and schedule computer processes for execution, perform memory management, provide file system, networking, I/O services, and provide a user interface, such as a graphical user interface (“GUI”), among other things.

Network

In the embodiment of FIG. 6, the data analysis system 100 is electronically coupled to a network 160, which comprises one or more of a LAN, WAN, and/or the Internet, for example, via a wired, wireless, or combination of wired and wireless, communication link. The network 160 communicates with various computing devices and/or other electronic devices via wired or wireless communication links.

According to FIG. 6, in some embodiments information may be provided to or accessed by the data analysis system 100 over the network 160 from one or more external data source(s) 172. The local data store 170 and/or external data source(s) 172 may include one or more internal and/or external data sources. In some embodiments, one or more of the databases or data sources may be implemented using a relational database, such as Sybase, Oracle, CodeBase, MySQL, and Microsoft® SQL Server as well as other types of databases such as, for example, a flat file database, an entity-relationship database, and object-oriented database, and/or a record-based database

Other Embodiments

Each of the processes, methods, and algorithms described in the preceding sections may be embodied in, and fully or partially automated by, code modules executed by one or more computer systems or computer processors comprising computer hardware. The code modules may be stored on any type of non-transitory computer-readable medium or computer storage device, such as hard drives, solid state memory, optical disc, and/or the like. The systems and modules may also be transmitted as generated data signals (for example, as part of a carrier wave or other analog or digital propagated signal) on a variety of computer-readable transmission mediums, including wireless-based and wired/cable-based mediums, and may take a variety of forms (for example, as part of a single or multiplexed analog signal, or as multiple discrete digital packets or frames). The processes and algorithms may be implemented partially or wholly in application-specific circuitry. The results of the disclosed processes and process steps may be stored, persistently or otherwise, in any type of non-transitory computer storage such as, for example, volatile or non-volatile storage.

In general, the word “module,” as used herein, refers to logic embodied in hardware or firmware, or to a collection of software instructions, possibly having entry and exit points, written in a programming language, such as, for example, Java, Lua, C or C++. A software module may be compiled and linked into an executable program, installed in a dynamic link library, or may be written in an interpreted programming language such as, for example, BASIC, Perl, or Python. It will be appreciated that software modules may be callable from other modules or from themselves, and/or may be invoked in response to detected events or interrupts. Software modules configured for execution on computing devices may be provided on a computer readable medium, such as a compact disc, digital video disc, flash drive, or any other tangible medium. Such software code may be stored, partially or fully, on a memory device of the executing computing device, such as the programming notebook system 100, for execution by the computing device. Software instructions may be embedded in firmware, such as an EPROM. It will be further appreciated that hardware modules may be comprised of connected logic units, such as gates and flip-flops, and/or may be comprised of programmable units, such as programmable gate arrays or processors. The modules described herein are preferably implemented as software modules, but may be represented in hardware or firmware. Generally, the modules described herein refer to logical modules that may be combined with other modules or divided into sub-modules despite their physical organization or storage.

The various features and processes described above may be used independently of one another, or may be combined in various ways. All possible combinations and subcombinations are intended to fall within the scope of this disclosure. In addition, certain method or process blocks may be omitted in some implementations. The methods and processes described herein are also not limited to any particular sequence, and the blocks or states relating thereto can be performed in other sequences that are appropriate. For example, described blocks or states may be performed in an order other than that specifically disclosed, or multiple blocks or states may be combined in a single block or state. The example blocks or states may be performed in serial, in parallel, or in some other manner. Blocks or states may be added to or removed from the disclosed example embodiments. The example systems and components described herein may be configured differently than described. For example, elements may be added to, removed from, or rearranged compared to the disclosed example embodiments.

Conditional language used herein, such as, among others, “can,” “could,” “might,” “may,” “for example,” and the like, unless specifically stated otherwise, or otherwise understood within the context as used, is generally intended to convey that certain embodiments include, while other embodiments do not include, certain features, elements and/or steps. Thus, such conditional language is not generally intended to imply that features, elements and/or steps are in any way required for one or more embodiments or that one or more embodiments necessarily include logic for deciding, with or without author input or prompting, whether these features, elements and/or steps are included or are to be performed in any particular embodiment. The terms “comprising,” “including,” “having,” and the like are synonymous and are used inclusively, in an open-ended fashion, and do not exclude additional elements, features, acts, operations, and so forth. Also, the term “or” is used in its inclusive sense (and not in its exclusive sense) so that when used, for example, to connect a list of elements, the term “or” means one, some, or all of the elements in the list. Conjunctive language such as the phrase “at least one of X, Y and Z,” unless specifically stated otherwise, is otherwise understood with the context as used in general to convey that an item, term, etc. may be either X, Y, or Z. Thus, such conjunctive language is not generally intended to imply that certain embodiments require at least one of X, at least one of Y, and at least one of Z to each be present.

While certain example embodiments have been described, these embodiments have been presented by way of example only, and are not intended to limit the scope of the disclosure. Thus, nothing in the foregoing description is intended to imply that any particular element, feature, characteristic, step, module, or block is necessary or indispensable. Indeed, the novel methods and systems described herein may be embodied in a variety of other forms; furthermore, various omissions, substitutions, and changes in the form of the methods and systems described herein may be made without departing from the spirit of the inventions disclosed herein. The accompanying claims and their equivalents are intended to cover such forms or modifications as would fall within the scope and spirit of certain of the inventions disclosed herein.

Any process descriptions, elements, or blocks in the flow diagrams described herein and/or depicted in the attached figures should be understood as potentially representing modules, segments, or portions of code which include one or more executable instructions for implementing specific logical functions or steps in the process. Alternate implementations are included within the scope of the embodiments described herein in which elements or functions may be deleted, executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those skilled in the art.

All of the methods and processes described above may be embodied in, and partially or fully automated via, software code modules executed by one or more general purpose computers. For example, the methods described herein may be performed by the programming notebook system 100 and/or any other suitable computing device. The methods may be executed on the computing devices in response to execution of software instructions or other executable code read from a tangible computer readable medium. A tangible computer readable medium is a data storage device that can store data that is readable by a computer system. Examples of computer readable mediums include read-only memory, random-access memory, other volatile or non-volatile memory devices, CD-ROMs, magnetic tape, flash drives, and optical data storage devices.

It should be emphasized that many variations and modifications may be made to the above-described embodiments, the elements of which are to be understood as being among other acceptable examples. All such modifications and variations are intended to be included herein within the scope of this disclosure. The foregoing description details certain embodiments of the invention. It will be appreciated, however, that no matter how detailed the foregoing appears in text, the invention can be practiced in many ways. As is also stated above, it should be noted that the use of particular terminology when describing certain features or aspects of the invention should not be taken to imply that the terminology is being re-defined herein to be restricted to including any specific characteristics of the features or aspects of the invention with which that terminology is associated.

Claims

1. A computer system comprising: one or more computer processors configured to execute software code; anda non-transitory storage medium storing software code that, when executed by the one or more computer processors causes the computer system to: in response to receiving one or more user inputs defining sequential analyses to perform on a data set, generate an analysis path visualization,wherein the analysis path visualization includes a hierarchical plurality of nodes corresponding to the sequential analyses,wherein the nodes of the hierarchical plurality of nodes are visually joined to indicate hierarchical relationships among the nodes, andwherein the nodes of the hierarchical plurality of nodes comprise icons representative of types of the respective sequential analyses.

2. The computer system of claim 1, wherein at least one or more nodes of the hierarchical plurality of nodes further comprise indications of user inputs defining the sequential analyses.

3. The computer system of claim 2, wherein the indications comprise query information provided by a user.

4. The computer system of claim 1, wherein the software code, in response to execution by the one or more computer processors, further causes the computer system to: in response to a first user input selecting a node of the hierarchical plurality of nodes, generate a data visualization based on the data set and an analysis associated with the node.

5. The computer system of claim 4, wherein the software code, in response to execution by the one or more computer processors, further causes the computer system to: cause display of the data path visualization and the data visualization adjacent to one another.

6. The computer system of claim 5, wherein the software code, in response to execution by the one or more computer processors, further causes the computer system to: in response to a second user input, update the data path visualization to display all paths associated with the sequential analyses.

7. The computer system of claim 1, wherein the hierarchical plurality of nodes includes key nodes associated with the sequential analyses and not other non-key nodes associated with the sequential analyses.

8. A computer-implemented method comprising: by one or more computer processors executing software code: in response to receiving one or more user inputs defining sequential analyses to perform on a data set, generating an analysis path visualization,wherein the analysis path visualization includes a hierarchical plurality of nodes corresponding to the sequential analyses,wherein the nodes of the hierarchical plurality of nodes are visually joined to indicate hierarchical relationships among the nodes, andwherein the nodes of the hierarchical plurality of nodes comprise icons representative of types of the respective sequential analyses.

9. The computer-implemented method of claim 8, wherein at least one or more nodes of the hierarchical plurality of nodes further comprise indications of user inputs defining the sequential analyses.

10. The computer-implemented method of claim 9, wherein the indications comprise query information provided by a user.

11. The computer-implemented method of claim 8 further comprising: by the one or more computer processors executing software code: in response to a first user input selecting a node of the hierarchical plurality of nodes, generating a data visualization based on the data set and an analysis associated with the node.

12. The computer-implemented method of claim 11 further comprising: by the one or more computer processors executing software code: causing display of the data path visualization and the data visualization adjacent to one another.

13. The computer-implemented method of claim 12 further comprising: by the one or more computer processors executing software code: in response to a second user input, updating the data path visualization to display all paths associated with the sequential analyses.

14. The computer-implemented method of claim 8, wherein the hierarchical plurality of nodes includes key nodes associated with the sequential analyses and not other non-key nodes associated with the sequential analyses.

15. A computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by one or more processors to cause the one or more processors to: in response to receiving one or more user inputs defining sequential analyses to perform on a data set, generate an analysis path visualization,wherein the analysis path visualization includes a hierarchical plurality of nodes corresponding to the sequential analyses,wherein the nodes of the hierarchical plurality of nodes are visually joined to indicate hierarchical relationships among the nodes, andwherein the nodes of the hierarchical plurality of nodes comprise icons representative of types of the respective sequential analyses.