Process Discovery Groups

This section contains Process Discovery Group settings. Groups specified here logically combine computers with installed Process Discovery Agents, such as finacial_dept, hr_dept, sales_dept, or others. When you deploy the Process Discovery Agents, specify a name of the group created here.

To create a new group, click Process Discovery Groups on the left menu, specify necessary parameters on the Create new group tab and then click Save. A new tab with the specified Process Discovery Group name appears in the Process Discovery Groups section and you are automatically navigated to the newly created group tab.

If you are editing settings in the existing group, click Save to save the settings after editing.

Option

Description

Group settings

Specify general group parameters.

  • Group name: Specify a new or edit the existing group name.

  • Record: Enables action recording for this group.

  • Analyze: Enables action analysis for this group.

Recording settings

Application ignore list: This is a comma-separated case-insensitive list of process names of the programs that Process Discovery ignores when recording and processing data. A program process name can be viewed in the Windows Task Manager. As ignored programs you can specify messengers that usually are not part of the work process. Specify this list to increase the Agent and Analyzer performance and reliability of discovered processes by removing extra "noise" from data.

Agent database settings

Specify database parameters for Agents to access and store data.

  • Address: Database server name or IP address.

  • Port: Specify the port to access the database.

  • Schema: Name of schema.

  • Type: Specifies the type of database. This parameter cannot be changed.

  • User name: Account name to access the database by the Agents in this group.

  • Password: Account password to access the database by the Agents in this group.

  • Test button: Checks the connection to the database.

Agent database provisioning

Enter credentials to create the database schema and grant access to the user specified in Database settings. These options are available in the existing group.

Note Database provisioning buttons are disabled until you save your settings.
  • Database administrator name: Specify the name of the administrator account to access the database.

  • Database administrator password: Specify a password for the administrator account to secure it.

  • Create schema button: Creates the schema in the database using the specified credentials. If the schema exists, you are prompted to overwrite the existing schema or cancel.

  • Grant user access button: Grants necessary database access rights to the user specified in Database settings. If the user exists, you are prompted to recreate the existing user or cancel. If you recreate the user, the previous user access rights and other user settings are lost.

Analysis settings

Specify Analyzer run settings for this group. See the "Analysis settings explained" section below for more information.

  • Min count of process instances: Set the minimum number of similar sequences of user actions to form a process.

  • Min action count in process instance: Set the minimum number of user actions in process instances to discover.

  • Max count of sequential noise actions in process instance: Set the maximum number of consecutive user actions unrelated to a process instance.

  • Max inactivity interval in process instance (minutes): Set a maximum user inactivity period in discovered process instances.

  • Max application count in process instance: Set a maximum number of different applications in a process instance.

  • Blur screenshots: If this option is selected, Analyzer makes the text on screen shots unrecognizable for the Insight dashboard views. The screen shots in the Agent database remain intact. We strongly recommend selecting this option if the screen shots can contain sensitive data, such as personal information, passwords, credit cards numbers, and so on.

Advanced recording settings

Specify advanced recording settings for this group.

  • Non-UIA applications: Enter a comma-separated list of process names of the programs for which Process Discovery uses ISA (Intelligent Screen Automation) instead of UI Automation API. By default, the field contains names of some communication programs, Java programs, all browsers except Chrome, and others that show more reliable results when using ISA. Because Kofax RPA implements native UI automation support for the Chrome browser, we recommend using Chrome as the default browser in your environment.

  • Agent SQLite DB max size (MB): Set the maximum size of the data stored locally on the computer where an Agent is running. The Agent stores collected information locally if connection to the database cannot be established. If the stored data exceeds the limit, the Agent starts overwriting the oldest data. The default setting provides storage for two to three days of data collection. Increase the size limit if you often experience long network outage.

  • Min UIA tree immutability duration (ms): This is a time the Agent waits before recording a program tree after the tree stopped changing. The lower the timeout, the faster the Agent records the tree and the better results you can get. If programs on the user computers become slow in response, increase this number. Do not change the default setting without thorough consideration.

Advanced analysis settings

Specify advanced Analysis output settings for this group. Do not alter the default values of the following settings without thorough consideration.

  • Max compute score distance (0-100): Sets the proximity score for comparing user actions based on the number of actions in groups.

  • Min action block score (0-100): Provides a resemblance score threshold for combining adjacent action blocks. Decreasing the number makes Analyzer more tolerant and combine more blocks. Increasing the number makes Analyzer more precise in combining action blocks.

  • Min strong pair score (0-100): Sets user action resemblance score to combine actions into pairs while detecting processes. If you decrease the value, you can get more discovered processes with more false detections and vice versa.

  • Max distance between process instances (0-100): Sets the minimum resemblance score of the detected process instances to combine them to clusters. The lesser the specified number, the more processes are considered similar and combined, but you can get more unalike processes combined in clusters. If you increase the value, more clusters are created with less process instances in them.

  • Actions in blocks for comparison: Sets the maximum number of user actions in blocks used for comparing and combining those blocks. If you decrease the value, Analyzer becomes more tolerant to "noise" actions and detects more processes, but you get shorter processes and might miss some longer ones. If you increase the number, Analyzer may start combining more blocks with unalike actions.

  • Max ISA X position difference (px): Sets the maximum element offset in pixels on the x-axis until marking it as a different element when detecting similar actions. This setting is applied to processes detected using ISA technology. If you increase this value, more actions are considered similar, which may lead to more false process detections and vice versa.

  • Max ISA Y position difference (px): Sets the maximum element offset in pixels on the y-axis until marking it as a different element when detecting similar events. This setting is applied to processes detected using ISA technology. If you increase this value, more actions are considered similar, which may lead to more false process detections and vice versa.

Analysis settings explained

See the Process Discovery glossary for terms used in this section.

  • Min count of process instances: This number specifies how many times a sequence of similar actions occur in the collected data to mark that sequence as a process. The number must be consistent with the amount of data you collect (number of people, duration of the data collection process, and so on). If the number of users is large and you spent considerable time collecting the data, you can increase the quality of the analysis by specifying a larger number to cut off short repeated actions that are not beneficial to automate, to shrink the size of the report, and to decrease the analysis time. Before you change the number, it is good to have a general idea of the work in the selected department. For data collected during two weeks or more in a department with 20 people or more, you can start with a number 15 and then increase or decrease the number depending on the discovered processes in the report. We recommend analyzing the same data with different values in this and other properties to refine the report.

  • Min action count in process instance: This property specifies a number of unique user actions (such as a mouse click, entering text in a text field, copying text, and so on) that the Analyzer must discover and mark as a process. If you want to discover processes that have different number of unique actions, such as five and fifteen, you must specify the lower number (5) in this field.

  • Max count of sequential noise actions in process instance: This is an additional setting that helps the Analyzer to consider some user actions unrelated to the process instance while discovering this instance and cutting a sequence of actions into process instances. This setting allows to discover process instances even if during a usual course of action a person is distracted by erroneous clicks or mistypes, important instant messages, other tasks withing the same application, and the like. Note that a process instance cannot contain more than the specified number of consecutive noise actions in it. For example, if while performing a usual sequence of actions a person started doing something else and exceeds the number of noise actions specified in this field, the Analyzer considers this process instance finished. Increasing the number in this setting may lead to discovering larger process instances that stuck together, while decreasing the number leads to fragmentation of the process instances.

  • Max inactivity interval in process instance (minutes): This setting specifies the maximum duration of a person's inactivity at the computer, that is, when a person is away from computer. A process cannot be discovered if user inactivity time exceeds the specified value. The value you specify must be a little over the inactivity periods in the processes you want to discover.

  • Max application count in process instance: This property must contain the maximum number of applications involved in a process instance. Before changing the default value, consider the number of applications your employees use when performing their usual tasks. The large number of applications may lead to unreliable data.