Download presentation
Presentation is loading. Please wait.
Published byShannon Bridges Modified over 9 years ago
1
Review of Building Intelligent.NET Applications Stu Egli Andre Inistotov Frenny Saldana Kate Styers Nishant Zinzuwadia MSE 614 February 26, 2008
2
The Book Assigned the book Building Intelligent.NET Applications by Sara Morgan Rea The book uses Microsoft Visual Studio and the.NET environment to develop Artificial Intelligence Applications Examples provided in the book include –An application built using the Speech Application SDK (Software Development Kit) –A voice-only telephony example –Another application using the Speech Application SDK –A prediction application that utilizes SQL Servers Analysis Services –A distributive schedule application using SQL Servers Analysis Services –A rule-based application –An agent based application
3
Why Use.NET for Artificial Intelligence Applications? Microsoft is one of the largest software vendors The author believes that Microsoft will be making software for many years to come Microsoft Research will continue to support the development of technologies like Artificial Intelligence to support continued use of Microsoft products The author utilizes the Speech Application SDK (Software Development Kit)
4
Minimum System Requirement for a.NET Artificial Intelligence Application Processor -- Pentium III class (600 MHz or higher) RAM -- 256 MB or higher Operating Systems –Windows 2000, service pack 3 –Windows 2003 Server –Windows XP Professional, service pack 1 Hard disk space -- 950 MB Soundcard -- Speakers are optional Microphone -- Headset with close-fitting microphone (for speech processing applications)
5
Software Requirements for a.NET Artificial Intelligence Application
6
Languages Provided with Visual Studio Visual Basic.NET C# (C sharp) J# (J sharp -- Microsoft’s version of Java [special add-on to Visual Studio])
7
AI Technologies explored in Book Intelligent Agents Usually has 3 – 4 of the following characteristics –Functions independently –Reacts to its environment –Able to communicate –Personalized –Able to learn Ex. Windows file update system by multiple remote monitoring agents to keep the user system updated by checking xml file that lists all the available file using Microsoft’s FileSystemWatcher class & transferring all the possible updates through Microsoft’s Background Intelligent Transfer Service.
8
AI Technologies explored in Book (cont.) Data Mining Also known as Data Warehousing enables use of huge amount data available and make useful predictions. Ex. Analysis Services Create a mining model includes: Creating a Database Naming the Data Source Using mining model wizard to create actual model Creating a training dataset to make prediction results
9
AI Technologies explored in Book (cont.) Rules – based System Capturing knowledge of expert and encapsulating into set of rules, programming it into computer code checking its reliability against set of inputs 4 phases used to apply all potential rules: Conditioning matching – returns a true value & utilizes System.Reflection namespace to invoke condition methods. Priority Evaluation – Identifies first phase based on name,default flag & priority, accepts its and passes generated result to next phase. Application of Results – orders rules based on priority, executes each result method and create log indicating rules applied. Adjustment – Monitors ResultLog & ResultLogDetail tables, adjusts rules when application determines matching question & condition, invokes a result method or encounters an error.
10
AI Technologies explored in Book (cont.) Speech Processing The speech engine recognizes what the user says by applying predefined grammar rules Speech Application SDK once installed creates new speech application for Speech Web Application Prompt Editor managing Prompt Database which contains prerecorded phrases used by user. If not Text-to-Speech will speak Telephony Application Simulator and Speech Debugging Console are used in developing and debugging voice-only application. Call Viewer and Speech Application Report, provided my Microsoft Speech Server allows to analyze and report on call event data.
11
Other Areas of AI and Microsoft Research Fuzzy Logic Game AI Genetic Programming Natural Language Processing Neural Networks Machine Learning Robotics Visual Studio.NET is a wonderful tool for creating powerful distributed Applications. The upcoming versions will be even more powerful. They will be based on some exciting technologies coming out of Microsoft Research.
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.