Roger D. Meicenheimer Botany Replacement of textbooks with interactive CD ROM expert systems $10,006 May 15, 1996 - July 7, 1996 1. Specific Aims and Objectives Types of Botanical Knowledge Required by the Public. Botanists, regardless of their subdiscipline, are generally asked two types of questions about plants by people, who are nonbotanists, when they learn they are talking with a professional botanist: 1. What is the identity of plants growing on their property? 2. What is wrong with specific plants they are attempting to grow? These questions indicate the general unfamiliarily that many people have with regard to plant identification and cultivation and the prevelant need that these people have for expert advice on these subjects. The need to replace textbooks. Our new "paperless" society now uses more paper per capita than before the widespread utilization of computers. This increase demand drives a continuing increase in paper prices, which is reflected in higher textbook prices. It is unreasonable to expect students to continue to buy and use textbooks and incur additional expenses associated with computer technology. As the Superintendent of the Illinois State School System recently said, "In fact, it could be argued that the computer ought to be replacing the textbook as a fundamental basic learning tool." (Pearson, 1995). Expert Systems. The "tricorder" featured in the popular "Star Trek" books, television series, and motion pictures, is a futuristic device by which, humans, interacting with a portable input device, are able to identify and evaluate unknown objects and situations. Tricorders are still science fiction constructs, but "expert systems" that provide similar capabilities for specific categories of unknown objects or situations are becoming reality (Pennisi, 1994). Expert systems will dramatically alter the way in which we interact with one another and our environment. The need for experts. The development of expert systems is dependent upon a body of experts, who can provide the information, or data bases, necessary for that technology to perform the expected functions. Our department trains close to 200 students per year in the field of plant taxonomy. We need to build upon their extensive training by incorporating the development of plant identification expert systems into their curriculum. Specific objectives. The objectives of this proposal are to: 1. Enriched the educational experience of students by involving them in the creation of an interactive CD ROM expert system which will replace two textbooks that are currently used in Dendrology (Botany 205). 2. Improve instructional effectiveness by engaging undergraduates in the creation of this expert system which will require utilization of their newly/simultaneously acquired knowledge of trees and wood into a format that is both self reinforcing and beneficial to future students. 3. Test the effectiveness of this method of learning compared to current pedagogic practices. BACKGROUND AND CONTEXT OF PROJECT Current objectives and structure of Dendrology. The objective of Dendrology (Botany 205) is to provide the student with an intensive and broad experience with respect to knowledge of trees for majors in botany and other biological sciences. This experience includes identification and distribution of native and introduced trees, characteristics and utilization of their woods, and an introduction to forestry practice. The four credit hour course is structured as two lectures followed by two, two hour laboratories per week. Tree identification is stressed during the first half of the course and wood identification is stressed during the second half of the course. Current textbooks. There are three required textbooks for Dendrology. The Textbook of Dendrology (Harlow, et al., 1995) supplements the lecture component of the course. Guide to the Woody Plants of the Tri-State Area (Blackwell, 1976) is a classical dichotomous key that is used for tree identification by students in the first half of the laboratory component of the course. Computer-Aided Wood Identification (Wheeler, et al., 1986) is a supplemental text to a multiple entry computer program for wood identification, entitled "Guess" that is used by students in the second half of the laboratory component of the course. The total cost of these texts is currently $96. Current pedagogy. The pedagogical approach taken in teaching students to identify trees and their woods are currently very different from one another. In the case of tree identification, the student must successful proceed through a very structured sequence of dichotomous decision making steps to successfully reach the species name of an unknown specimen (See Appendix A). In the case of wood identification the student must successful complete a data entry sheet on anatomical features, enter these data into a computer, which then searches for matches of their entries against a stored data base on species specific wood features (See Appendix B). There currently is no one data base that contains both tree and wood features of the same species. Nor, is there currently a unified procedure by which a student can use to identify an unknown tree and its wood. Potential of an expert system. The technology currently exists with which to develop an expert system for both tree and wood identification thereby providing students with a modern unified procedure to identify unknown specimens. One such "expert system", entitled XIDAUTH, has been used successfully to develop expert systems for the identification of agriculturally important weed species and insect pests for the USDA Animal and Plant Health Inspection Service (Pennisi, 1994). With the XIDAUTH program, the student inputs in a freely interactive manner, the characteristics that are most obvious to them for a given specimen. The stored data base is searched after each characteristic entry, progressively eliminating the size of the potential species list, until a positive identification is made for an unknown. Once an unknown has been identified, full text description and graphical images can be accessed to verify the match between the species and the unknown specimen. Other information about the identified specimen is also readily accessible from within the XIDAUTH program. Current state of Dendrology expert system. In the summer of 1995, I utilized some of my departmental instructional funds to purchase the XIDAUTH program to begin developing the Dendrology expert system. Currently I am working on the development of a data base and menu structure that contains species characteristics of trees and their woods, with the ultimate intent of providing Dendrology students access to this expert system via a CD ROM. A poll of current Dendrology students indicated that they would be amenable to purchasing such a CD ROM, provided the cost was around $60 and that it would replace the two textbooks used in the laboratory component of the course. III. PROPOSED PROJECT Methodology and Procedures The structure of the Dendrology Expert System (DES) will be subdivided into five major components of species specific characteristics (Appendix C). The data base will be constructed using various diverse sources of information pertaining to each of the components that are readily accessible to students. The menu structure will be supplemented with the creation of text and digital images pertaining to on-line help with regard to various character states used in the menus. Text and digital images for each species within the data base will be created. The data base will be further supplemented with regard to incorporation of the following additional information on each species: 1. Species distribution maps, 2. Silviculture information on each species, 3. Information on the commercial use of the species. Undergraduate students who have successfully completed Dendrology will be trained in digital image processing techniques to help with the acquisition of these images. The same students will also help in development of the data base entries and accessory information contained on the beta version of Dendrology expert system. Finally, these students will serve as initial users of the DES for initial debugging purposes. This procedure is a lengthy iterative process which will take at least eight months to complete. Once the beta version of the DES is complete, it will be tested by Dendrology students prior to development of the alpha version. III. EVALUATION AND ASSESSMENT OF PROJECT RESULTS Evaluation. The beta version of the DES will be made available to a group of Dendrology students whom will use this in place of the two laboratory textbooks. The control group will be a group of students whom will use the current textbooks. The effectiveness of the CD ROM as textbook replacements will be assessed by comparing the mean scores of the overall performance of these groups of students with one another. Assessment. In addition to the quantitative evaluation of the effectiveness of the CD ROM replacing textbooks on student performance, each student who participated in this project by using the DES will be asked to complete a questionnaire assessing its utility as a pedogical device and to provide written commentary on the menu structure and data base images. The responses to these questionnaires will be used in the development of the alpha version of the DES, which will be incorporated into the Dendrology curriculum the following year. V. DISSEMINATION OF RESULTS ASSESSMENT OF POTENTIAL IMPACT AND DISSEMINATION PLANS Two hundred Miami University students enroll in Dendrology (BOT 205), Plant Taxonomy (BOT 202), Field Botany (BOT 155), and Tropical Flora of the Bahamas (BOT 470) each year. Each of these courses contains a component pertaining to identification of trees and involving these students in the development of the DES will help prepare these students in modern approaches to plant identification. Students and faculty involved in courses in plant identification at other universities potentially will be interested in the DES. In addition, there currently exists a text market for tree identification books written for the general public. The DES would presumable have appeal to a subset of that audience, whom would have the necessary hardware to access the information contained on it. BUDGET LTEP BOTANY DEPT Personnel Faculty Faculty Summer Stipend .............. $ 4,000 Students Undergraduate Student Wages ($4.50/hr X 320 hrs) ................. $ 720 $ 720 Equipment A. 90MHZ Pentium Dell Dimension XPS P90 W/3X CD ROM computer with 16 MB RAM, 256K Writeback Cache, 1 GB Hard disk, 3.5 Floppy disk drives, Del VS17 Monitor (17" CRT, non-interlaced 64 bit #9GXE64 Pro PCI Local Bus Video Card w/ 2 MB VRAM), NEC 3X CD-ROM, 450KB/Sec., Dell Mouse, 101-Key Keyboard, Dos 6.2, Windows 3.1, Sound Blaster 16 Audio Card and Peavey 200 speakers ........ $4,100 B. Pinnacle RCD-202 Recordable CD Drive with RCD-PC software, Adaptec Compatible SCSI Host Adapter and drivers ..................... 2,700 C. Toshiba Satellite Pro 75MHz Intel486 Notebook computer with 500MB hard disk, DS CDROM drive, 1 MB video RAM, 4 MB RAM 3 X $2,500 ea. ................... 7,500 B. Videodiscs and CD ROM disks ..... 100 ... $ 200 TOTAL (COMPLETE PROJECT = $15,941) ........ $25,020 .... $5,020 BUDGET JUSTIFICATION Faculty and undergraduate summer salaries are necessary for the completion of the summer objectives. Since I will be devoting 100% of my effort to the proposed project instead of my research program, I will undertake the proposed development of the DES only if adequate compensation for my invested time is provided. The indicated hardware is necessary for the development of the CD ROM and to conduct the assessment of the DES by students.