Analyzing Cases¶

The Cases View¶

While the Map view (Analyzing Process Maps) gives you an understanding about the process flows, and the Statistics view (Analyzing Statistics) provides you with detailed performance metrics about your process, the Cases view actually goes down to the individual case level and shows you the raw data.

You get to the Cases view by simply changing to the Cases tab as shown in Figure 1.

Figure 1: The Cases view in Disco.

To be able to inspect individual cases is important, because you will need to verify your findings and see concrete examples particularly for “strange” behavior that you will most likely discover in your process analysis. Almost always you will find things that are hard to believe until you have drilled down to an individual example case, noted down the case number, and verified that this is indeed what happened in the operational system.

Furthermore, looking at individual cases with their history and all their attributes can give you additional context (like a comment field) that sometimes explains why something happened. Finally, the ability to drill down to individual cases is important to be able to act on your analysis. For example, if you have found deviations from the described process, or violations of an important business rule, you may want to get a list of these cases and talk to the people involved in them to provide additional training.

The Cases view consists of the following areas:

Complete log (1)

Select the Complete log item on the left to see a list of all cases in your data set in the second column to the right, see (3) in Figure 1.

Individual variants (2)

Alternatively, you can select an individual variant, whereas a variant is a specific sequence of activities. Only cases that follow the same activity sequence are then displayed in the list of cases in the second column (3). The variants are sorted by their frequency.

For example, in Figure 1 the third most frequent variant has been selected. There are 63 cases (accounting for 10.36% of the whole log) following the activity sequence Create Purchase Requisition -> Analyze Purchase. One of these cases (case 538) has been selected and is displayed. Refer to Inspecting Variants to learn more about variants and how to take advantage of them in Disco.

List of cases (3)

A list of cases IDs for either the complete log (1) or the selected variant (2).

Individual case (4)

In the main window you can see further details about the currently selected case from (3). Refer to Individual Cases for further details about the detailed case view.

Search (7)

You can search for specific case IDs or attribute values. Read Search to see how searching cases works.

Filtering (8)

The log filter controls for the current data set can be accessed from each of the analysis views. Filters are really important drill into specific aspects of your process and to focus your analysis. Read Filtering for detailed information on how filtering works in Disco.

Copy, Remove, and Export data set (9)

Data sets can be copied, deleted, and exported right from the current analysis view. Read Copying Data Sets for further details on copying and deleting data sets. Process maps can be exported, for example, as a PDF file. The export functionality of Disco is explained in detail in the Export reference in Export.

Inspecting Variants¶

Variants are an integral part of the process analysis. In Disco, a variant is a specific sequence of activities. You can see it as one path from the beginning to the very end of the process. In the process map (Analyzing Process Maps), an overview of the process flow between activities is shown for all cases together. A variant is then one “run” through this process from the start to the stop symbol, where also loops are unfolded.

Usually, a large portion of cases in your data set is following just a few variants. For example, in the purchasing process shown in Figure 1 the top five most frequent variants cover the process flows of ca. 50% of all cases while there are 98 different variants in total.

It is useful to look at the distribution of cases over variants in your data set to:

• Know what the most common activity sequences are. For example, in Figure 1 one can see that the third most frequent variant is the activity sequence Create Purchase Requisition -> Analyze Purchase. In total 63 cases (10.36% of all cases) follow this particular pattern from beginning to the end of the process. From a process improvement perspective this seems strange. Why are so many cases stopped? Should the purchasing guidelines be updated to clarify what people can buy and what not?
• See how much variation there is. Besides inspecting the most dominant sequences, it is also useful to know how many variants are there in the first place. For example, popular methodologies like Lean Six Sigma focus on reducing variation to improve quality and process efficiency at the same time. In fact, there is often much more variation in processes than people are aware of. Knowing just how much variation there is (how many variants in total) helps to get a sense of how streamlined your process is.
• Simplify your process map. In situations where you want to focus on the mainstream behavior, you can view the process map and statistics for just the most frequent variants. To do this, you can use the Variation Filter as shown in Figure 2. After applying the filter, the process map for just the mainstream process flows are shown in Figure 3.

Figure 2: The Variation Filter lets you select a portion of more frequent variants (mainstream behavior) or less frequent variants (exceptional behavior).

Figure 3: The process map of the purchasing process—covering 50% of the cases in the data set—after filtering for just the Top 5 most frequent variants.

Note that there are certain types of processes, which do not exhibit many common activity patterns. For example, in a hospital the diagnosis and treatment process for almost each patient is unique. These kind of processes are often called “unstructured” and the variants and the Variation Filter are not of much help in these situations. Instead, you can use the simplification controls of the Map view (Analyzing Process Maps) to understand and analyze the process flows.

Individual Cases¶

You can inspect individual cases by selecting one of the case IDs in the list of cases, see (3) in Figure 1. Information about the currently selected case is shown in the main window, see (4) in Figure 1.

In this case view, you have some overview statistics about the case in the upper part, and a detailed history view in the lower part of the screen (Figure 1):

Time frame of case (5)

This chart shows you where in the log the selected case occurs. The blue chart in the background is the Active cases over time chart from the global statistics view (Overview Statistics). The time frame of the currently displayed case is highlighted in a bright red, so you can easily see where in the context of the overall log timeline (for example, more in the beginning or towards the end) it occurred.

Next to the time frame chart you find the following statistics:

• Events. The number of events (i.e., the length of the history) for this particular case. For example, the case 538 displayed in Figure 1 has 2 events, while case 1031 from Figure 4 has 17 events in total.
• Start. The start time for the earliest activity in the history for this case.
• Duration. The time between the earliest and the latest timestamp (total duration).
• Active time (only available if you have start and end timestamps for your activities). Percentage of time from the total duration of the case that was spent on performing activities (active time).

Figure 4: The table view shows the history of the case in a compact tabular form with all attributes that were included during import.

Graph and Table view (6)

You can switch between a Graph and a Table view for inspecting the concrete history for the case. The Graph view is shown in Figure 1 and only displays the activity sequence plus time and resource information. Additional data attributes are hidden.

If you want to see the data attribute values, or if you have longer cases and find yourself scrolling around in the Graph view of the case too much, you can switch to the Table view for a more compact representation of the case history including all the attributes that were imported (see Figure 4).

Search¶

The search function of the Cases view allows you to quickly look for cases that contain a specific text fragment either in the case name or in one of the attributes.

For example, imagine that you are analyzing the purchasing process from the demo example and you quickly want to inspect a case where a dispute occurred. Instead of adding a filter to drill down to cases containing a dispute activity (see Filtering Activities from the Process Map), you can simply search for dispute in the search field of the Cases view, see (7) in Figure 1 as shown in Figure 5.

Figure 5: Searching for a text snippet gives you a list of all matching cases and highlights the matching events in the case.

You can not only search for text snippets from the activity names but from all attributes in the data set. The matching events are highlighted in orange to show you where in the case your search has given a result.

Figure 6: The search for case IDs helps you to find specific cases in your data set.

Finally, you can also search for specific case IDs. Just type the name of the case in the search field and Disco gives you the case history you are looking for. An example case search is shown in Figure 6. This can be particularly useful if during your analysis you are comparing the data from the operational system and the analyzed data in Disco to verify your findings.