Web search as an interactive learning environment for graduation projects.

This paper presents a novel concept--a Web search environment as an interactive learning environment for graduation projects. The paper addresses the question of an appropriate learning environment for writing graduation projects in universities: Can it be defined? What should it be? How should it look? How and when should it be used? The premise presented is that one of the best learning environments for writing original, high academic level graduation projects already exists and is widely used--the Web search capability and its environment. A survey of this subject and its analysis is presented.

**********

INTRODUCTION

This paper presents a novel concept, another view of a Web search--the use of Web search as an interactive learning environment (ILE) for writing an academic graduation project (GP). Writing GPs is a compulsory task for students receiving a bachelor's degree in Israeli and in most other universities and colleges. Students must write several mini-projects or seminar papers to earn their degree (the number of projects varies from two to six). The project or seminar paper should be an original paper based on current research and development in the area of the student's choice (or more correctly the area of interest to the instructor (Zviel-Girshin, 2003)).

One of the questions that is of immense importance for achieving the educational goals of writing a GP is "Can a good learning environment be designed for such a project?"

This paper analyzes the process of writing GPs in the past, present, and future. It claims that one of the best and omnipresent learning environments for this purpose is Web search engines and the Web browser environment.

THE PAST

Mini-projects or seminar courses are academic courses taken mostly during the student's last year in a university or college. An instructor or a team of instructors selects subjects and proposes subjects about which students might write. The aim is research, self-study, and a deeper understanding of the subjects. Instructors usually recommend some relevant articles and books for initial reading. At the end of the term, each student writes an original GP of relatively high academic level. The project should present the selected subject, explaining the problems, solutions, and innovations in that area of research. In case of a mini-project in computer science or engineering, students sometimes develop applications such as software, hardware, or electronic circuitry.

During the term students are encouraged to find more materials related to the selected subject and enrich their own knowledge of it. In the past, before the World Wide Web, information retrieval usually took place in the library. The knowledge of the librarian and library catalogs were used. According to statistical data, the search and information retrieval took from 2-3 days to 2-3 weeks (Zviel-Girshin, 2003).

In the past, the information retrieval process was library-centered. The library database was the basic and commonly used tool for data retrieval. Databases were not necessarily computerized and most used different formats of information encoding, labeling, and classification. Consequently, students had to study special query formats and searching techniques; they also had to use the help of others to search the library databases. Freshmen in Israel took a compulsory course of several hours about library use. The course was given by one of the library employees. The librarian explained the basic library structure, including how to search for publications in specific areas of expertise, how to match a publication number with its location in the library, and how to retrieve a book. In libraries that have a computerized index, its format and how to use it was explained to the students. An additional problem with libraries and their databases was the selection of the publications. Most of the libraries kept a great number of journals and publications about popular, required topics. However, they had only a small number of journals and books in more specific and narrow areas of expertise. Therefore, students who sought materials in an "unpopular" area often could not find them and had to place a special order and wait for it to arrive. The LP pattern of the learning process and the data retrieval mechanism in the past are described below. In the beginning, the instructor proposed a list of subjects. After discussions with the instructor about potential subjects, a student chose one and received a list of publications. Then the student performed library searches and acquired as many as possible of the required materials. Reading, analyzing, and rereading the literature was the next vital phase of the research; this resulted in the actual writing of the graduation paper. The instructor could do several revisions before the GP was submitted. Computers were used indirectly for doing database searches and for typing and printing the final papers.

The LE should be a set of teaching and learning tools designed to enhance the student's learning experience. Keeping in mind the described learning pattern it seems that an appropriate tool for enhancing the GP learning process did not exist. The LE for GP project research was very primitive and non-interactive. Therefore, it would not be too great an exaggeration to claim that until recently students did not have any LE of value, certainly not any computerized or interactive LE.

THE PRESENT

Different studies (Zviel-Girshin, 2003) have shown that the pattern of the LE remains almost the same. The subject of the GP is still selected by the instructor; the instructor still recommends some basic publications for initial reading. However, students themselves perform most of the search for information sources. The data retrieval process today is performed using library databases, the Internet, and Web search engines.

During the last decade, the Internet has become a useful tool for seeking information from all over the world. Library databases are updated more slowly than Web libraries. Many journals publish articles or abstracts of the articles on the Web. Some journals are e-journals and publish their materials on-line only. Numerous authors publish an abstract or a full-text article on their homepages. State-of-the-art discoveries, developments, and research results are published immediately and can be retrieved from all over the world without delay.

The pattern of the learning process and knowledge retrieval today is described below. As in the past, in the first phase, the instructor advises the student and supplies the primary list of publications. In the next stage, the student performs Web and library searches, analyses the potential sources, and searches for new facts, materials, and publications. Then the student writes the final paper. However, on the information highway students can access and retrieve new information that may be "unknown" to the instructor. They can retrieve up-to-the-minute information; most of this "unknown" data is found using Web searches.

Once again, the questions "Can a good LE for writing a GP be proposed?" and "What should it be?" are raised. To answer those questions we should analyze the tools used by students during their research and try to understand the functional meaning of an LE and the interactions involved.

LEARNING ENVIRONMENT FOR GRADUATION PROJECTS

For many years, instructors of mini-projects and seminar courses tried to find a good ILE for writing a GP, an environment that would give enough broad and new information in theoretical, practical, and technical topics, an environment that would force students to think, to investigate, and to research. In today's world, traditional library-based searches are not current enough. The information placed in the libraries is not sufficient to satisfy the global village environment. Some of the topics can to a degree be researched using libraries only (e.g., literature, linguistics, history), but most disciplines require more detailed, technical, or recent information. Therefore, the best and most current LE for writing a GP is the browser environment and Web-based search engines.

To test the above premise, a special research project was conducted. A written survey was performed. Three groups of students from different universities and colleges were asked to answer the survey questions (Table 1).

One group was from a university with a large library and two groups from colleges with significantly smaller libraries. Altogether 82 cross-disciplinary students answered the survey. In this survey, students were asked some basic questions about their mini-project or seminar course project; these included questions about data retrieval tools and techniques used and the time they devoted to the project. The survey questionnaire can be found in the Appendix.

Survey Analysis and Results

The analyzed data produced the following results:

* 52% of the students had 24-hour access to the Internet

* 93% of the students used a Web search engine during the project

* 67% of the students who performed Web searches found relevant information that they used in their final papers

* 87% of all the students still used a traditional library search

* All 100% of the university students (i. e., students who had a large library at their disposal) were in this group.

Additional analysis of the languages used during the search showed a strong correlation between finding no relevant information and the language used (Table 2).

Ninety percent of the students who did not find relevant information used the Hebrew language during their search. This strong relationship between language used and success is by no means incidental. The reality is that in languages other than English, and particularly Hebrew, significantly fewer articles are published or indexed by search engines. Therefore, the conclusion is to advise students to use the English language because it is the "nonofficial" language of the Internet.

A more specific analysis of project subjects, query strings, and retrieval of relevant material showed the basic problems to be:

* Incorrect query string or lack of ability to structure a query string that exactly defines the required subject

* Language of the query strings (i.e., some of the searches were performed in Hebrew)

* Language preferences of the browser interface and search engine

* Poor translation of basic terms from Hebrew into English and from English into Hebrew

* Irrelevant information or "junk results" as the result of too short a query string.

Time related questions provided the following data:

* 73% of the students who used Web-based searching used it between three and fourteen times

* 74% of the students used Web search engines once a week or once in a fortnight (Figure 1)

* 68% of the students devoted between 2-3 weeks or 1-2 months to the project (note: the semester length in Israeli universities is 14-17 weeks) (Figure 2)

The retrieved information was mostly downloaded by students to their home computers, read, and saved for further use.

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

* In 58% of the cases, the same search was performed more than once, using more and more refined queries; these additional queries were formulated based on the results of previous queries of the same question (Figure 3)

No correlation or dependency between search length (i.e., time invested in the search), the finding of relevant materials, and time devoted to the project was found. The Web search patterns of the students in the university (i.e., with large library) were found to be the same as those of the students at the college (except for wider library use of the university students).

In future research, a correlation between time spent in retrieving material, writing the paper, and the final grade should also be measured. Students who took part in the survey were cross-disciplinary students; therefore, it will be interesting to check if there is any relationship between a student's major and the time spent in Web searching, retrieval of relevant material, and the success of the search and project writing.

Is it Really an Interactive Learning Environment?

ILE refers to the presence of a "computer" in LE (Reeves, 1997). ILE is a Web-based environment that supports structured interaction between a community of learners (VirtualSchool, 2003). Interactive environment refers to the environment that accepts input from users (Webopedia, 2003). ILE should provide a learning environment in which the time, the place, and the form are convenient for learners.

Web searching can be, and usually is, performed at home or at another location convenient to the researcher; it can be performed at any time. A Web browser usually supplies materials in a format that is comfortable for reading (e.g., html, doc, or simple txt file). Therefore, the browser environment and Web-based search engines are very convenient for learners and can be used as an ILE for writing a GP.

The environment itself only supplies the information. It is learner's task to read and analyze it. Some of the retrieved materials will require additional self-study and research. The additional research can be done in an interactive way using a computer, Web browser, and input from the user:

* Previously retrieved materials are downloaded to computer and analyzed by the user

* As the result of this analysis, the user refines or changes a previously used query string or formulates a new one

* The Web search engine produces a new result that is displayed by the browser.

This additional data retrieval, understanding, and research are performed at a place and a time that is convenient to the learner; the learner's task is to analyze this new information.

THE FUTURE

The future development of Web search engines and browser environments leaves no doubt that this combination will be one of the most frequently used tools for current data retrieval. Therefore, it will be used by almost everybody (Chen, Dal Busco, Garrett, Sinha, 2000), and especially by students who have been encouraged to search for information in some field of expertise. At present, only half of the students have 24-hour Internet service or other network access. However, it is reasonable to assume that in near future almost all the students will have 24-hour, direct access from the student's PC to Web search services. Even now, many articles are published or indexed on the Web by their authors; in near future, most articles will be published on the Web immediately. The majority of journals have an on-line copy available in which at least article abstracts are available to all. More and more journals have become exclusively electronic and more and more researchers prefer to use an electronic version of a journal rather than a hardcopy version (Rusch-Feja, Siebeky, 1999). The same trend will continue in the future. In the near future, the browser will understand all formats of electronic document encoding (e.g., .doc, .pdf, .ps, .zip) or one general format for Web publishing (e.g., XML) will be defined. This will solve the problem of lack of standardization (Rusch-Feja, Siebeky, 1999). For that reason, the browser environment will be very comfortable for viewing and reading written publications and with the addition of Web search engines will be the ILE of choice for GPs.

CONCLUSIONS

Browser environments and Web search engines must be studied as an ILE in universities. The ILE stimulates learning and encourages students to think and to research. Interactions with this LE are by queries. The LE receives a query as an input and provides the result of the search as output. After reading the result, the student can refine the search by adding, deleting, updating, or improving some information from the previous search. After downloading, reading, and understanding the retrieved information, the student can refine the query once again or completely change the direction of the search. The learning process continues until the student is satisfied or has enough information to finish the paper.

A more profound analysis of the survey described above showed that it is advisable to give a brief course about the use of Web search engines. The best time to give this course should be at the beginning of a student's freshman year. During this course, students should be taught a brief introduction to the interfaces of the browser and Web search engines, what a query string is, and how to write a well-structured query string. The students must understand that query strings should not be too short--otherwise too many pages are retrieved, most of which are irrelevant. Query strings should also not be too long--otherwise the student receives no data at all. Students must understand that it is advisable to use English as the query language in view of the fact that most of the articles published on the Internet are written in English. However, since most of the students write their GP during their last academic year they should have no problems with the English translation of the basic field-specific terms. (This may vary according to the country, language used, and project topic).

The answer to the question about a good ILE for writing a GP is clear-browser and Web-based searches are the most suitable LE for this purpose. The data retrieval can be performed in a time and place and using a format convenient for the student. This environment interacts with the user and provides broad theoretical, practical, and technical information in all areas of research. This combination of Web search and browser is widely used by students who find it very comfortable, functional, and useful.

Appendix

Survey Questionnaire (translated from Hebrew)

Do you have 24-hour access to the Internet?                     yes  no
Did you use Web-based search engines during your project?       yes  no
Did you find relevant information?                              yes  no
Did you use a traditional library search?                       yes  no
If you did not find any relevant information on the Web,
try to specify your project's subject
If you did not find any relevant information, try to specify
query strings you used
How many times did you use the search engine? (Number of
times--twice in the same day may be counted as 2 times if the
waiting period was more than 3-5 hours).
How often did you use the search engine? (Daily usage, weekly
usage, once in 2 weeks, monthly usage).
How many days were devoted to the project?
Did you use additional Web searches as a result of
previously downloaded information and its analysis?             yes  no
Did you use English in your Web search?                         yes  no
Did you use Hebrew in your Web search?                          yes  no

Table 1 Research Population

                    University  College A  College B

Number of students  38          20         24

Table 2 Search Success vs. Search Language

Search Language              Successful search  Failed search  Total

Hebrew                        8%                30%             38%
English and other languages  59%                 3%             62%
Total                        67%                33%            100%

no   42%
yes  58%

Figure 3. Feedback-driven Search

Note: Table made from pie chart.

Acknowledgment 

This paper is based on research partially supported by a research grant from Magbit Canada and Jordan Valley College.

References

Chen, M., Dal Busco, J., Garrett, K., & Sinha, A. (2000). Search engine usage. Retrieved November 20, 2002, from University of California, Berkeley, School of Information Management and Systems, Survey Project Requirement for IS 271, Web site: http://www.sims.berkeley.edu/~sinha/teaching/Infosys271_2000/SearchEngine/

Reeves, T. C. (1997). The scope and standards of the journal of interactive learning research. Journal of Interactive Learning Research, Educational Technology Review, 7, 5-8.

Rusch-Feja, D., & Siebeky, U. (1999). Evaluation of usage and acceptance of electronic journals: Results of an electronic survey of Max Planck Society researchers including usage statistics from Elsevier, Springer and Academic Press. D-Lib Magazine, 5(10).

Zviel-Girshin, R. (2003). Topics for the academic project. JVC Technical Reports TR-2003. virtualschool.edu (2003). Retrieved August 8, 2003, from http://www.virtualschool.edu/ile/-ILE

Webopedia.com (2003). Retrieved July 20, 2003, from http://www.webopedia.com/

RINA ZVIEL-GIRSHIN

Open University Israel, Jordan Valley College, Israel

[e-mail]

NATHAN ROSENBERG

Advanced Research Center, Jordan Valley College, Israel

[e-mail]