This system has some certain limits of itself, namely

• The links embeded in the articles are added by human experts, which is a tedious and time consuming task. In our future work, it should resort to automatic links generation. Although automatic links generation may add some meaningless links, even bad links into the articles, it can really find some hidden links which cannot be explicitly found by human beings. I believe that automatic link generation is a promising direction in hypertext system research.
• The efficiency of this system is not satisfying. The index is built for each node of an article, thus a lot of ad-hoc searching and computations occur during users' queries. New methods should be developed to improve the system in terms of processing efficiency.
• This system is in a closed domain and it does not interact with materials outside, especially on the Internet. This problem should be overcomed by extending the links to outer links, not only inner links.
• Since this system is developed based on an Encyclopedia, it is not necessarily fit for scholarly article set. For scholarly paper collections, more indicative materials can be utilized, such as abstracts, conclusions and references. In fact, references and bibliography can be considered as natural links to other works. More work should be done on automatic links among scholarly articles. Some initial work has been started in Citeceer, which is a computer science paper search engine and links papers together using their references and bibliographies.