November 09, 2016

BDB Platform

New Features:
  1. Web Single Sign-On with SAML has been implemented to authenticate with a client’s SAML identification provider.
  2. Data Service
    1. Dedicated Connection Engine: Implemented for each data service.
  3. System Admin.
    1. Sessions: A new administrative module has been added to display active users
    2. Monitoring: Capture and store the information in data store.
  4. Installation Package: Installation script has been implemented for Linux servers.
  5. Download option has been provided for Data Service.
  6. Reactivate option has been provided for expired users.
  7. Status option has been provided as a filter value in the 'User Management' module.
  8. Added Data Store connector for Dashboard Designer.
Enhancements:
  1. Active MQ Advanced
    1. Dead Letter Channel: Controls actions such as whether to propagate the thrown exception to the caller (the handled option), and where to direct the (failed) exchange.
    2. Active MQ Persistence: Preserves data to recover undelivered messages in the case of a server failure.
  2. Platform Core
    1. ODATA Enhancement: ODATA version 4 has been implemented to introduce security mechanisms.
  3. Schedule Data Fetch
    1. a. Data Store has been improved to load data into the BDB data cache faster
    1. Query Service Engine can serve data to the dashboards faster
    1. Improved adhoc data processes have been implemented on data aggregation and caching
  4. The 'Share' option and 'Latest Schedule Status' notification window are provided for Data Store Service
Please Note:
  • Data Store connector is currently in Beta.

Business Story

New Features:
  1. KPI tile has been added to the ‘View’ menu row to display the sum of the quantity
  2. The 'Share View' option has been added to share a view across the story
Enhancements:
  1. Elastic Search has been introduced to improve data querying performance by 10x
  2. Adjustable chart grid is provided in the Storyboard to customize view size
  3. Filter Enhancements:
    1. Multiple filter values can be selected
    2. Global Filter notification bar has been removed
    3. 'Apply' option has been provided on the filter panel
    4. The view specific filter icon has been provided separately in Story Designer
    5. Gobal filter will be applied only to views where the filter values are used as dimensions
  4. Business views have been enhanced
  5. 'Show Legend' option has been enhanced

Dashboard Designer

New Features:
  1. Charts:
    1. Added Circumplex chart
    2. Repeater properties for the Line and Bar charts to display the values by individual categories
    3. Dynamic range and bar indicators in the grid components to set a color for the Alert option
    4. Chevron shape in the bar and group bar chart (to show progress of projects when the available data is in numbers format and not in date format)
    5. Clustered view added to the Group Bar chart
    6. '100%' view added to the TimeLine chart
    7. Curve line for best fit line added to the Scatter Plot chart
    8. Property option added to the Score Card to display all the nodes in close/open state when grid loads
    9. Properties option in the Radio Button to set labels in horizontal/vertical orientation
  2. Designer:
      1. Added HTML editor to the Text Box component to write a aligned and formatted description
      2. BDB logo added to the Designer screen
      3. 'Save As' option provided in the design area
      4. Context Menu options added to set equal space vertically or horizontally between multiple components
      5. Double click action (on the dashboard list) to launch dashboard in the designer
      6. 'Delete' has been provided for groups
Enhancements:
  1. Charts:
    1. Message is displayed when dashboard is previewed and it does not have any data assigned to it
    2. Smooth loader to prevent flickering while loading data
  2. Designer:
      1. Object Browser window upgraded to allow reordering of z-index of components by dragging and dropping
      2. Action labels have been renamed in publish dashboard model for clarity
      3. Search feature enhanced in component Attribute window
      4. Full screen view option added in the component script window
      5. Component can be selected via right click

Predictive Workbench:

New Features:
  1. Multiple workflows are now supported
  2. Components for the following Apache Spark Features(Alpha release):
    1. Data Preparation
      1. Data Split - Splits data set into train and test based on train and test ratio
      2. Data Type Definition - Identifies/classifies data type in the following formats:
        1. Integer
        2. Double
        3. String
    2. Data Transformation
      1. R Formula - Produces a vector column of features and a double or string column of label
      2. String Indexer - Encodes a string column of labels to a column of label indices ordered by label frequencies
    3. Algorithm - Spark Naive Bayes (with Validation) has been provided to introduce Spark Machine Learning
    4. Validation
      1. Train validation for splitting the dataset based on train ratio
      2. Cross validation for splitting the dataset based on number of folds
      3. User can set evaluation metrics as Binary or Multi-class
      4. Users can view the conditional probability chart using the 'Visualization' tab (if the Naive Bayes model has been completed)
    5. Performance - Users can compare up to three workflows using the performance component. Users can set evaluation metrics as Binary or Multi-class
    6. Save and Read Model - Users can train a model and apply it to test data. Models can be saved and reused
    7. The 'Force Start' option has been provided for Spark jobs, if the spark request queue becomes full
    8. The 'Clear Cache' option is provided to clear cached data
    9. Users can track live status of Spark jobs (with a view option) by enabling the log option
    10. 'Predictive Spark App Settings' option provided in the Administration module
  3. 'Permissions' for Predictive Workbench has been added under the User Management module to assign access permission to user groups
  4. 'Share With' and 'Copy to' options are provided to share workflow/model with a user or group
Enhancements:
  1. Cassandra Data Writer has been provided with a provision to view 'Replication factor'
  2. 'Stop' button has been added to the Progress Bar
  3. 'Delete Connection' option has been provided for inter-node connections
  4. Minimize/Maximize buttons added to enhance space management in the workflow pane
  5. 'Search' option provided for the entire tree node
  6. Due to browser constraints, the data set that can be viewed in the table/chart has been limited to 50,000 cells
  7. Custom R Script –
    1. For obtaining summary, the user needs to specify summary() explicitly in the return statement. Ex return(list(out=output,model=summary(kmeans_model)))
    2. In 'Property view definition' settings – Support for multiple data types (string, integer, and double) for text box control
Please Note:
  • Spark and R workflows work independently. When using Spark workflows, users can't use R components and vice versa. (i.e. If a user drops Spark components to the workspace, s/he can't drop R components on to the same workspace)
  • Spark workflow only supports Cassandra Data Source/Data Writer
  • Spark components do not support 'Run Till Here' feature
  • R workflows do not support 'Clear Cache' and ‘Refresh’ features
  • Spark workflows do not support scheduling
  • Before a user can save a Spark model, the work flow must be run
  • Due to browser constraints, the data set that can be viewed in the table/chart has been limited to 50,000 cells
  • Spark String Indexer, Spark R Formula, Spark performance, Spark Apply Model, Spark Read, and Save model are currently supported only for Spark Naïve Bayes.

Survey

New Features:
  1. Analyze Result section:
    1. 'Filter Rules' function added
    2. Functionality to create a new view has been added
    3. 'Show Rule' functionality has been added
  2. Survey Templates have been added
Enhancements:
  1. The preview page has been provided with various preview icons:
    1. Desktop
    2. iPad
    3. Mobile
  2. Scheduler has been implemented in the DataMart section

Social Media Browser

Enhancements:
  1. Advertisements in Facebook posts/comments eliminated using machine learning

Sentiment Analyzer

Enhancements:
  1. Some sets of 'Stop Words' have been removed
Please Note:
  • Sentiment Analyzer : - NLP (Natural Language Processing) of Sentiment Analyzer is currently at the Beta stage. Our language experts are continually upgrading and improving the NLP dictionary and algorithms. It can be used internally, but users may experience defects and usability issues.
General Recommendations:
Product Description
Dashboard Designer For best performance while using ‘Paging Grid’, records should be kept under 5000.
Dashboard Designer For best performance while using Excel/CSV as connector, Excel/CSV file size should be under 3 MB size.
Dashboard Designer Excel/CSV data should be structured as same as database table.
Dashboard Designer Due to limited browser memory, A chart can draw up to 150K data points more data can lead to browser crash.
Dashboard Designer Country/State names should be properly mentioned, while using map component.
Dashboard Designer More than 60 chart/grid components in dashboard will make DOM heavy and leads to browser failure.
Dashboard Designer For best performance while viewing a dashboard, the latest version of browsers should be used.
Business Story A server with 32GB RAM and 4 Core CPU are required to configure Business Story application.
Business Story Elastic search sever should be allocated minimum 16GB.
Business Story Approximately 20 user concurrences are supported (on this mentioned configuration) to a story which handle ~12 million records.
Business Story For best performance while scheduling Data store use ‘Batch query’ option effectively.
Business Story Usage of ‘*’ and ‘limit’ are not recommended in SQL select statement of Data Store (E.g. select top 100 from employee, select * from employee) .
Business Story Segregation of fields into Dimensions, Measures, and Time Groups should be done properly .

Known Limitations:
#ID Description
23546 Platform- To list the newly created query service, ‘Refresh’ button must be clicked.
23531 Platform-Textbox filter is displayed as a drop-down menu .
23530 Platform-Report does not display data while selecting all the filter value.
23529 Platform-Partial data will be displayed when all filter values are selected in multiselect .
23528 Platform-Context menu permissions are getting checked while editing a user group, even if it is not selected to be checked.
23512 Platform-File selected using ‘Preferences’ is displayed as blank .
23484 Platform-Administration - ‘Edit’ option is provided as an icon for R-Server configuration details under Predictive settings .
23483 Platform-Selection of ‘General Settings’ options cannot be saved in Audit trial .
23470 Platform- Scheduled old Data store is working with invalid DB details in Data connection .
23356 Platform-Admin instances are displayed twice in the session management system and users cannot apply ‘Kill Session’ option on one system admin session.
23343 Platform-Scheduler status message is not proper.
23339 Platform-Data Management Module - ‘Cube’ is displayed instead of ‘Data Store’ in the pop-up messages (while saving a data store and for scheduler success/failure messages).
23338 Platform-Deleted Data Store instance is present in AMQ .
23311 Platform-Username and password in Oracle connection can be copied to other connection .
23270 Platform-Old group permissions are not reflected in the new code push.
23242 Platform-While using ‘Share With’ option, group all users are listed in exclude .
23239 Platform-Users cannot download the shared query service .
23189 Platform-While creating a connector from the service level (JUNIT), it is created with a special character .
22047 Platform-Connections are not released from DB .
22036 Platform-Login page placeholder should be generic .
21701 Platform- ‘Download’ and ‘Print’ options are not required for reports in iPad.
20791 Platform- ‘Re-publish’/ ‘Publish as New’ option is not shown for system documents.
20772 Platform-Non-mandatory (optional) filters for clarity service are not fetching the data.
20466 Predictive Workbench- Limiting number of rows to 50,000 cells
23424 Predictive Workbench- Test button to validate given credentials by user in spark settings.
23355 Predictive Workbench- y axis values are not proper in ROC curve .
23354 Predictive Workbench- Multiple charts do not plot in ROC and PR Curve .
23150 Predictive Workbench- Visualization for the Market Basket Analysis does not cover the entire page.
23099 Predictive Workbench- Scientific values are not displayed in tooltip of the scatterplot chart.
22025 Predictive Workbench- The UI gets stuck, if users click on ‘Live Job Status’ option when status of that process is in progress, but execution is completed.
19561 Predictive Workbench- Server is not able to take additional requests, if request is more than 300 or CPU utilization is high in R.
16032 Predictive Workbench- No intimation is provided to users, if server reached out of memory during any process.
16016 Predictive Workbench- Query service does not retain complete milli second value while passing timestamp or date having milliseconds..
15813 Predictive Workbench- Scheduler generates an error while writing data with date. The Error message comes like, 'Data truncation: Incorrect date time value: '0017-06-29 00:00:00' for column 'DateKey' at row 1'.
13899 Predictive Workbench- All the warnings from R are not displayed, only last values are shown (Only last values are displayed from R, not all).
12724 Predictive Workbench- No intimation is provided to users, if application server is down and process is running.
11513 Predictive Workbench- The Time Series chart will not plot, if data is in exponential form.
11454 Predictive Workbench- User should be restricted to view a maximum number of rules in the Market Basket Analysis.
22698 Survey- Not able to update/edit the selected year in the Date/Time question type
23165 Dashboard Designer- Alert comparison value equal to '0' cannot be filled properly for the bar shaped indicator provided under data grid.
23147 Dashboard Designer- ‘Width’ property within Dataset properties cannot work in data grid when provided less than 10 px.
20553 Dashboard Designer- Page unresponsive message is displayed in ‘Paging Grid’ while previewing dashboard that contains data more than 6K rows.
23536 Dashboard Designer- Headers in data grids also should be wrapped when ‘text-wrap’ property is set to true.
23510 Dashboard Designer- Indicator property container of grids is not scrollable. Some Properties cannot be accessed in small screen machine
20384 Dashboard Designer- The Pie chart shrinks in size when ‘Show Percentage’ label property is enabled and chart height is less then 300px
23192 Dashboard Designer- Line curve in the Time series chart is not persistence at the top for some continuous and equal data points.
22710 Dashboard Designer- Scatter Plot data set does not show changed name of the X-field in Tooltip chart .
21762 Dashboard Designer- Charts that have small radius of data points (such as Line chart and Area chart) are difficult to drill down in iPad .
21757 Dashboard Designer- Display for Tooltip chart is delayed by some milliseconds in iPad .
23480 Dashboard Designer- Legend text overlaps with the next legend, if the font size is reduced in iPad.
21758 Dashboard Designer- Tooltip for the Spider chart is not displayed in iPad/it is hard to click on the data points in iPad.
17123 Dashboard Designer- Slider in the Time Series chart cannot be dragged smoothly in iPad .
18087 Dashboard Designer- Users may not be allowed to type in the filter value text box within Page grid in iPad.
19062 Dashboard Designer- Series is displayed in chart when reload filter option is applied, but shows unchecked in legends.
19996 Dashboard Designer- Precision property can be applied to 0 value of Y axis for most of the charts
20183 Dashboard Designer- Connector panel in the Designer may keep on loading (it can be due to a network issue) .
16451 Dashboard Designer- ‘Undo’ feature in script window reverts script written in other components to the script panel.
18987 Dashboard Designer- ‘Undo’ feature may not work properly for the chart properties .
20560 Dashboard Designer- Horizontal/vertical scroll may appear by default in the dashboard, even if no changes have been made to any component.
18640 Dashboard Designer- Images and text are not clear in the exported pdf file.
18638 Dashboard Designer- Words are not clearly displayed for the Tag Cloud chart in the exported pdf file.
16875 Dashboard Designer- Exported PPT/PDF/PNG of SVG components like Time Series/ Map/ Sliders/ Graphics may have broken/improper images.
20512 Dashboard Designer- Some SVG images in dashboard are cropped for the earlier versions of dashboards .
17446 Dashboard Designer- V-slider component cannot be exported properly .
17447 Dashboard Designer- Component size shrinks in SVG image while exporting the image.