Multimedia Instructional Tutoring and Authoring System (MITAS) with 3D
MITAS is a Microsoft Windows-based tool for developing and presenting computer-based instruction. MITAS was originally developed under contract to the U.S. Army Research Institute. MITAS uses a unique approach to lesson development featuring:
- Multimedia
- 3-D graphical animation
- Speech recognition
- Natural language and dialogue generation
- Exercise development templates
- Lesson sequencing based on student performance
MITAS includes both an authoring module and a student module. Instructors can use the software to develop high-quality computer-mediated training lessons that incorporate the above technologies. MITAS supports various instructional approaches and can respond to student performance. Through the multi-media and natural language components of MITAS, new and truly innovative instructional strategies are possible. The visual lesson development environment eliminates the need for programming by the lesson developer.
Beta versions are available now. Information can now be obtained from MAAD_info@alionscience.com.
MITAS takes advantage of multimedia technology!
Authors can include video, sound, graphics, graphical animations and speech recognition in their lessons.
Supported multimedia formats:
Video: AVI, MOV, and MPG
Sound: WAV
Graphics: BMP
Graphical Animation: OpenGL-based
Speech Recognition: Entropic Cambridge Research Laboratory, Ltd. HAPI


Speech Recognition in MITAS
The MITAS 3-dimensional micro world exercise uses a corpus-based, continuous, speaker-independent speech recognition system. Components are built in for Arabic, Spanish and English. The speech recognition components were developed using the HTK application developed by Entropic Cambridge Research Laboratory.
Natural Language and Dialogue Generation in MITAS
Build natural conversation into your lessons!
MITAS includes a powerful natural language component for Arabic and Spanish. Central to the concept of an intelligent tutor is the ability of the tutor to understand and respond in the language being taught. Without a strong natural language environment, MITAS would be no better than any other computer-aided instructional environment. Our natural language processor (NLP) allows the author to build lesson structure around natural student responses instead of requiring highly specific responses. The NLP adds a syntactic component, a principled lexical, semantic component, and a knowledge base component to the application. This permits it to handle dialogue within a scenario presented to the student, thereby providing an environment where the student can become immersed in the use of the language. The NLP was developed by the University of Maryland.

Exercise Templates
How to build instruction with MITAS
One of the features of MITAS that makes it unique is the process of designing and building blocks of instruction. MITAS uses a flowcharting technique where a block of instruction is developed by drawing the links between the storyboard elements. When branching based on student responses is desired, the instructional developer simply draws the links between the storyboard elements and, then, double clicks on the branch point and enters the logic. Blocks of instruction can be developed hierarchically with branching logic between blocks of instruction as well as between storyboard elements.
The basic flow of instruction is developed in an easy to use lesson authoring environment as shown below.
Easy-to-use exercise development templates
The Authoring module was designed so that instructors without programming expertise can easily develop lessons. The lesson is defined using a graphical flowchart approach and each individual exercise is defined through the use of template screens.
The Authoring Module features the following capabilities:
- A palette of drawing tools that lets you develop exercise sets and exercises by drawing it on the screen.
- Exercise set, exercise, and sequencing description windows that you open by clicking on the network diagram.
- Easy to use exercise development templates for defining exercise parameters.
- An expression builder window that allows authors to create simple or complex mathematical and logical expressions to describe performance based sequencing options.
- Cut-and-paste capabilities for moving or copying exercise sets, exercises, or text within a lesson or between lessons.
- Automatic score data collection.
- The ability to reference a text, graphic, sound, and/or video file within an individual exercise.
- The ability to import/export developed lessons and all associated graphic, sound, and/or video files between computers Graphic animation.
- The ability to specify high and low priority syntactic and semantic errors.
- The ability to specify student feedback for syntactic and semantic errors.
All of the exercise templates were constructed so that exercise essential information is entered on the first screen (What is the question, what is the answer, how many tries does the student get, etc.). The second screen allows the author to specify information that may or may not be utilized depending on the exercise goals. For example, if an exercise will be used to teach reading comprehension, then the author will specify a text file to be used as the basis for the exercise. The system allows text and/or graphic and/or video files to be referred to in any exercise.
The exercise types that the tutor supports are as follows:
- Multiple Choice (Oral or textual prompt)
- Fill In the Blank
- Question and Answer (Oral or textual prompt)
- 4D Location (Oral or textual prompt, Single or Multiple locations, Ordered or unordered multiple identification)
- Sorting (Textual or Graphical)
- Menu-Created Sentences
- Cloze
- 3D Micro World
- Translation (English to Arabic/Spanish)
- Transcription (Arabic/Spanish to Arabic/Spanish)
- Pronunciation Practice
In addition to the above exercise types, the Author can also define introductory/tutorial exercise nodes and end of exercise set nodes.

Lesson Sequencing Based On Student Performance
MITAS includes the ability to measure student errors and revise instruction accordingly. This is accomplished using an expression builder that allows authors to create simple or complex mathematical and logical expressions to describe performance-based sequencing options.
Lesson sequencing can be fixed, random, student choice, or performance-based.
