Dr. David Raker Education Department Westfield State College Westfield, Ma 01086 413.572.5293 draker@javanet.com

Dr. Jamie Callan School of Computer Science Carnegie Mellon University Pittsburgh, Pa 15213 412.268.4525

Electronic information literacy is becoming increasingly more important in the current information age. Consequently, for young people to be successful in school and in work, they must have the ability to find, evaluate, and utilize electronic sources. Educators understanding this need continue to attempt to better integrate the use of electronic media (most often the Internet) with existing classroom curriculum. However, too often the reality of integration is confounded by concerns that students do not have the appropriate skills to locate and evaluate Internet based information or will too easily find inappropriate material. Thus, educators and parents must consistently try to balance their concerns with the clear utility of providing such a powerful tool for learning. Project K12-Search, therefore, is a collaborative effort that attempts to address many of these concerns and ultimately help schools better utilize the Internet. Specifically, Project K12-Search is in year two of a three year project and presently consists of both a set of student lessons that focus on the development of information literacy skills and a new type of Internet search engine interface that better reflects the needs and concerns of the K-12 environment.

Three main goals help guide our work:

• To develop strategies that will help students better understand what the Internet is and how it works.
• To explore guided learning opportunities that will help students better find, evaluate, and utilize Internet resources.
• To develop an Internet search engine that will better help students find and explore age and content appropriate materials throughout the Internet.

Lesson Plans

Project K12-Search consists of 14 lessons that can be used either as a unit or separately to help students in their learning of specific Internet related material. These lessons have been field-tested with elementary students, but could also be modified for students at the middle and secondary levels.

The lessons were developed to introduce a number of Internet and information literacy related topics. Some of these topics apply more specifically to younger students, but other students (of all ages) too can benefit. Although diverse in scope, each of the lessons has certain commonalties as well. For example, each lesson utilizes a dynamic, hands-on approach to learning. Also, both student and teacher evaluation is stressed throughout. Lastly, many of the lessons require little or no actual Internet access. This allows schools not yet "wired" to still introduce important topics to students who might have Internet access at home or through the library.

K12-Search Engine Interface

Too often the utility of Internet content must be weighed against the effort a child or teacher must expend to find material suitable for a particular lesson. Simply put, searching for information on the Internet can be time consuming and frustrating. One variable confounding a better match between the needs of educational institutions and the power of the Internet is the dependence on traditional Web search engines to find appropriate educational material. These Web search engines provide raw horsepower, but offer little assistance in guiding and harnessing this power. In other words, Web searches often suffer from high recall (many useful things found) and low precision (even more useless things found). Teachers, and especially children, do not have the time or the attention span to sift through large numbers of documents to find the few of interest.

K12-Search is developing software that helps bring the power of the Internet and Web search engines to the classroom in a form suitable for teachers and students. K12-Search is a new graphical user interface developed exclusively for K-12 schools and their specific needs. For example, one unique characteristic of the K12-Search interface is that retrieved documents are all listed with their corresponding reading level. This allows students to better choose documents based on their own reading level, instead of spending time trying to sift through material that is inappropriate for their use.

Broadly speaking, the K12-Search interface guides students through the process of formulating and describing their information needs. Modified versions of traditional Information Retrieval (IR) tools will transform the information need into a search query, utilizing knowledge of teacher objectives, student level, and subject matter. The intent is to let children express their information needs in English, have the software transform it into a complex structured query, and pass it off to the Web search engine. Returned results are organized automatically, using clustering and machine learning techniques. This is viewed as a 'second stage' of searching, done at the interface, in which results are organized and further filtering takes place.

Presentation

Specifically, the K12-Search presentation will include a demonstration of the various lesson plans and search attributes of the software illustrated above. In addition, particulars of how K12-Search software works and was developed will be described. Lastly, baseline data and anecdotal information based on experiences in schools will be shared.

Session attendees who have considered or are now using the Internet in school will have the opportunity to explore the new lesson plans, see K12-Search software in action, and examine how far the project has moved ahead in the second year.

Project K12-Search is a joint effort of the Language Technologies Institute at Carnegie Mellon University, the Westfield State College Education Department, the University of Massachusetts Center for Intelligent Information Retrieval (CIIR), the University of Massachusetts Library, Javanet, and Merriam-Webster Inc. K12-Search is funded by grants from the National Science Foundation, Library of Congress, and Department of Commerce (award IIS-9812358 and cooperative agreement EEC-9209623).

Appeared in The Proceedings of the 2000 National Educational Computing Conference, June 27, 2000, Atlanta, GA.