Introducing Microsoft Speech Server 2004 R2

With the release of Microsoft^® Speech Server (MSS) 2004 R2, Microsoft Corp. takes another step forward in its strategy to enable enterprises of all sizes to better serve existing customer needs, drive operational efficiencies, and create the next generation of customer service applications. MSS 2004 R2 is Microsoft's tool for deploying and managing distributed speech applications that combine Web technologies, speech-processing services, and telephony capabilities in a single, integrated, flexible, cost-effective platform.

MSS 2004 R2 enables large-scale enterprises and midsize organizations to reap the significant business value of speech technology by helping to reduce the industry's typically high cost of adoption, development, deployment, and long-term management of speech applications. Together with the Microsoft Speech Application SDK (SASDK) Version 1.1, building, deploying, and managing speech-enabled Web and telephony applications for a broad range of client devices is a more realistic goal than ever for organizations of all sizes.

Among the most anticipated new features of MSS 2004 R2 are the introduction of full support for North American Spanish and Canadian French and improved deployment flexibility.

Support for U.S. Spanish and Canadian French at No Additional Cost
MSS 2004 R2 enables you to deploy speech-enabled applications designed for Canadian French, U.S. Spanish, or U.S. English speaking users with the addition of MSS Language Pack Kits for Spanish (United States) and French (Canada). An in accord with Microsoft's aim to offer speech technology for the lowest total cost of ownership in the industry, MSS 2004 R2 provides support for these languages at no additional cost, including language-specific automatic speech recognition (ASR) and text-to-speech (TTS) engines, grammar libraries, tools and interface resources, documentation, and ASP.NET Application Speech Controls support files.
Improved Deployment Flexibility
MSS 2004 R2 introduces a new Enterprise Edition single-server deployment option designed especially to meet the needs of organizations requiring medium-sized speech deployments while significantly lowering purchase cost. This additional deployment option allows both Telephony Application Services (TAS) and Speech Engine Services (SES) components to be hosted on the same computer, and is ideal for deployments servicing between 25 and 48 ports, or for touchtone solutions supporting up to 96 ports on a single server. So whether your business requires a smaller standard single-server deployment, a larger enterprise single-server deployment, or an even larger distributed multiple-server deployment, MSS 2004 R2 has the flexibility to scale to your needs.

Added Features, Added Value

Microsoft Speech Server is already the industry's most flexible, most highly integrated speech application platform with the lowest total cost of ownership. MSS 2004 R2 adds significant value with the introduction of multilingual support and improved deployment flexibility. MSS 2004 R2 further introduces a number of other new value-adding features including:

Support for Call Analysis by Language
In addition to developing and deploying speech applications in new languages, the new language support features enable IT Professionals and Administrators to use MSS 2004 R2 analysis tools to:
- Filter queries by the language callers use.
- Parameterize reports by language.
- Select calls and other components such as QAs by language.
- Filter on event types during a database import for better management of the content and size of databases.
Upgraded ASR Engine Performance
The MSS 2004 R2 ASR engines for all languages deliver significant improvements in performance compared to the U.S. English ASR engine in MSS 2004, especially when loading and using large grammars. The main factors for this improved performance include:
- Grammar state caching optimization.
- Process recycling optimization.
- Decreased memory use.
- A higher degree of speech recognition accuracy.
Superior Reliability
While MSS 2004 already provides exception reliability, this is further improved in MSS 2004 R2 through the use of new, advanced asynchronous interprocess communication calls between processes such as Lobby, SES, and SES workers. This enables SES to recover automatically from error conditions such as remoting call loss and process deadlock. As a result, MSS 2004 R2 delivers superior interprocess call fault-tolerance, improved user experience, and a call service success rate in excess of 99%, even during peak capacity usage conditions.
Process Recycle Performance Optimizations
MSS 2004 R2 optimizes the management of issues surrounding process handling by implementing expanded use of process recycles. A process recycle shifts call handling from one process to another and can be triggered, for example, when the first process reaches its memory limit. The expanded use of process recycles by MSS 2004 R2 enhances reliability, scalability, and performance, especially at the extreme ends of usage capacity and application complexity. This enables you to service more users faster, with more reliability, and with fewer resources than previously.
Expedited Loading of Large Grammars from Saved State Information
Large speech application grammars place heavy demands on a speech platform's performance and scalability due to increased loading time and memory use. MSS 2004 R2 significantly minimizes this problem by enabling an ASR engine instance to save its grammar loading state onto disk. This saved state information can then be loaded into other ASR engine instances, thus expediting grammar loading and reducing overall memory use.
Extended Integration with Windows® Media Player and SQL Server Reporting Services
MSS 2004 R2 enhances your investment in Microsoft's development, Web, and data technologies and infrastructures by extending the integration of Log Analysis Tools with Windows Media Player and SQL Server™ Reporting Services (SSRS). Using Call Viewer, you can now use the Play Audio feature to play audio back through Windows Media Player (or another audio player), giving you greater control over caller audio analysis. And with extended integration with SSRS, SES now provides improved error reporting to the tuning tools, enabling developers and administrators to diagnose problems and minimize the impact to their system and schedule. MSS 2004 R2 implements enhanced error reporting for:
- Grammar compilation errors.
- Grammar load failures.
- Invalid SSML errors.

Now, more than ever…

Microsoft Speech Server 2004 R2 delivers more flexibility, better performance, greater reliability, enhanced integration, and more of the benefits of speech technologies at the lowest total cost of ownership. See for yourself how MSS 2004 R2 delivers more value for your enterprise's future, today.