Using Default Reports
Using Default Reports
Speech Application Reports provide a list of basic reports that analyze the logging data imported from event trace log (.etl) files to the reports database. These reports are based on Microsoft SQL Server Reporting Services.
To view Speech Application Reports, first create a reports database as described in the topics under Importing Log Files. Then, open Microsoft Internet Explorer and browse to https://ReportServer/Reports, where ReportServer is the server containing the reports database, and click on the report folder name that you specified when performing the import.
Understanding the Default Reports
The default reports are described in the following table.
Report |
Description |
|---|---|
Call Aggregates |
Provides total calls aggregated by various time periods: each hour in a day, each day in a week, each day in a month, each month in a year. For each time period the report provides statistical measures: minimum, maximum, average and standard deviation. |
Call Volume |
Displays call volume across a single span of time, plus providing comparative data on disconnection types, transferred calls, and call duration. |
Dialog Overview |
For the selected time period, displays data on QA use and QA recognition confidence scores, plus command use. |
Errors |
Displays summary data on call completion, a count of tasks missing TaskStart and TaskComplete events, and occurrence counts for error event types. |
Messages |
Provides counts of messages sent by the SALT LogMessage function, and a breakdown by message name and content. |
Prompts |
Provides counts for bargein and user-perceived latencies, prompt ending status, individual prompt totals, and comparison of text-to-speech versus recorded prompts. |
Server Statistics |
Provides data on latencies, channel and port usage, and answering time. |
Turn Analysis |
Displays data on QA usage per call, sums and percentages for QA History values, compares speech versus DTMF input, and reports on time spent on pages. |
Tasks |
Reports on task completion, duration and progress, and the count of QAs used for each task. |
For each report, the user can select the time span, called and calling device, and report format as well as selecting from a variety of export formats. In some reports it is possible to drill down to the Task and QA level.
To quickly change time periods or switch between reports, click on the links in the report heading.
Many reports contain multiple pages. To access earlier or later pages, use the Next Page and Previous Page buttons in the toolbar.
Deploying the Default Reports
For information on deploying Speech Application Reports, see the topic Reporting Services Deployment Guide in SQL Server Reporting Services documentation.
Configuring the Default Reports
Use the toolbar on each report to configure the data that is displayed.
Querying by Date
To select a range of dates to include in a report, in the SQL Server Reporting Services toolbar enter values in StartDate and EndDate.
Querying by Device
To select one or more devices to include in a report, in the SQL Server Reporting Services toolbar enter values in Called Device and Calling Device. Use the % symbol as a wild card representing one or more characters.
Note When using a leading % symbol with strings in the query parameters, the embedded quick links in the reports may not function correctly. In some cases the query parameter may be corrupted in the link's query string. The workaround in these cases is to use StartDate and EndDate in the toolbar, rather than using the quick links in the reports.
Querying by Language
To select calls or QASummary events by language, in the SQL Server Reporting Services toolbar select a language from the Language drop-down list box. Languages will appear in the Language drop-down list box individually or in pairs, semi-colon delimited. Paired languages are for calls in multi-language applications where one QA control is a language selector and the remaining QA controls in the call support the second language. The listed languages are the ones found in the Speech Application Reports database. In the Speech Application Reports database, the language setting is stored in the Message property of each QASummary event, and copied to the Languages column of the Calls table in the database.
There are two types of reports:
Call-centric reports: Call Aggregates, Call Volume, Server Statistics, Errors, Messages, Tasks.
QA-centric reports: Dialog Overview, Turn Analysis, Prompts.
For the QA-centric reports, there will always be one language listed. For call-centric reports there can be one or more languages listed because a call can span multiple languages, while a QA control cannot. The following table describes how selecting language in a query affects the data returned in the report.
Report |
Description |
|---|---|
Call Aggregates |
Returns only calls matching the selected language. |
Call Volume |
Returns only calls matching the selected language. |
Dialog Overview |
Returns only QASummary events matching the selected language. |
Errors |
Returns only calls and tasks matching the selected language. |
Messages |
Returns only messages from calls matching the selected language. |
Prompts |
Returns only prompts from QASummary events matching the selected language. |
Server Statistics |
Returns only calls matching the selected language. |
Turn Analysis |
Returns only QASummary events matching the selected language. |
Tasks |
Returns tasks only for calls matching the selected language. |
In Microsoft Speech Server applications, language is set on a per-page basis and is determined by the value of the xml:lang attribute of the HTML element. It is the responsibility of the application developer to ensure that the language attributes of prompts, grammars, controls and SALT elements referenced on a page conform to that page's language attribute.
Querying by Task Name
To query by a particular Task Name select one from the available list and click View Report.
Querying by QA Name
To query by a particular QA Name select one from the available list and click View Report.
Using Click-through to Drill Down in Graphed Data
In some reports, clicking data bars in selected graphs expands the display of that data.
Report |
Click |
|---|---|
CallVolume |
Calls Received Across Time |
Messages |
Count Across SALT Message Names |
ServerStatistics |
Maximum Number of Channels Simultaneously in Use |
Tasks |
Task Started |
Calculating User Perceived Latency
The Server Statistics report includes a table showing statistical measures of User Perceived Latency. In addition to QA response latencies, this measure also includes latencies experienced before a call is answered. Specifically, the User Perceived Latency table describes a time span including the following events.
UserPerceivedAnswerLatencyEvent
UserPerceivedDtmfLatencyEvent
UserPerceivedRecognitionLatencyEvent
Editing a Report Definition Language (RDL) File
The templates for each of the default reports are stored as .rdl files at this path: Program Files\Microsoft Speech Application SDK 1.1\SDKTools\Log Analysis\Reports.
RDL is an XML-based language that contains information about report design. Since the .rdl files are XML-formatted, they can be edited using Notepad or any XML editor, or using Visual Studio .NET 2003. The .rdl files are useful as templates for writing entirely new reports that will use the Speech Application Reports database; or for extending the default reports to do things that are more application specific, such as:
show the Tasks report only for those calls where the user hung up, or
on the ServerStatistics report, generating AnsweringTime/UPL tables only for those calls where two or more ports were simultaneously in use.
To extend one of the default reports edit the appropriate .rdl file and publish it to the reports server, as explained in the topics under Report Manager How-To in the MSDN Library. The logging data import process always uses the .rdl files located at the path Program Files\Microsoft Speech Application SDK 1.1\SDKTools\Log Analysis\Reports. The import task uploads any .rdl files in that directory to the specified folder on the report server, and associates any data sources titled "Speech Reports Datasource.rds" with the reports database that received the import.
To use a customized report on a one-time-only basis, edit the .rdl file and publish it to the report server. To add a customized report to the set of reports that always get uploaded when data is imported, add that .rdl file at the path Program Files\Microsoft Speech Application SDK 1.1\SDKTools\Log Analysis\Reports. Use caution when adding .rdl files to this folder, since if an .rdl file at this location is accidentally overwritten, it affects all subsequent imports.
The import task will not overwrite datasources or reports if they already exist. To replace a report or datasource that already exists, delete it via the SQL Server Reporting Services Report Manager.