Improvement of interactivity in a learning system for W.W.W.

为了补充科学教育的补充讲座，几年前在局域网上开发了一个学习系统。我们的项目旨在通过分析用户对学习系统的访问日志来更好地了解用户的学习情况。在项目的前期工作中，引入了PostgreSQL作为数据库管理系统（DBMS）来进行访问日志。在项目启动的第一年，开发了几个基于数据库管理系统的学习情况分析工具。通过使用这些工具，我们可以获得以下信息。1.各教材的访问频率2.访问频率沿着时间尺度的分布3.从本地站点的访问次数4.单个用户的访问次数和在线测试结果。例如，我们知道计算机学习室中配备的终端机使用最频繁，但所有访问次数中有23%来自我们大学之外。访问频率在一天中有3个高峰，第一个高峰在上午10点左右，第二个高峰在下午1点到4点之间，第三个高峰在晚上9点到11点之间，从而为分析访问日志奠定了技术基础。但由于教材数量少、种类少，目前很难将访问日志数据与学习情况联系起来。因此，我们的重点是丰富教材，积累更详细的知识访问日志数据。对于前一个目标，它是非常有成效的。利用JavaApplet开发了几种动画式、动态显示的教材。通过对访问日志的详细分析，发现在线测试和在线公告板是访问最频繁的两个站点。此外，当用户在网上测试中感到困难时，他们立即来到正确的网页，在许多情况下得到一些提示。因此，很明显，测试或练习可以成为学习的强烈动机。

In order to complement a supplementary lecture for science education, a learning system has been developed on the LAN several years ago. Our project aims at better understanding about user's learning situation by analyzing access log to the learning system. In a preceding work to the project, postgreSQL was introduced as a database management system (DBMS) to conduct access log. And in the starting year of the project several analyzing tools giving some information about learning situation were produced based on the DBMS. By use of these tools, we can get the following information. 1.Frequency of access to each teaching material 2.Distribution of access frequency along time scale 3.Access times from a local site 4.Access times and results of online test about individual user. For example, we know that terminal machines equipped in computer learning rooms are used most frequently, but 23% of all access times are from outside of our university. And frequency of access shows 3 peaks in a day, first one is around 10 a.m. and second is between 1 p.m. and 4 p.m. and the third is between 9 p.m. and 11 p.m. Thus, we are successful to establish a technical base to analyze access log. But unfortunately due to lack of quantity and variety of teaching materials, it is hard to relate the access log data with learning situation at this moment. So, we put emphasis on enrichment of teaching materials and accumulation of further detailed knowledge about access log data. Concerning with the former target, it was very fruitful. Several teaching materials of animation type and with dynamic display were developed by use of Java Applet. From detailed analysis of access log, it is shown that online test and online bulletin board are accessed most frequently. And further, when users feel some difficulties in the online test, immediately they come to right web page to get some hint in many eases. Therefore, it becomes clear that test or exercise can be a strong motive to learn.