回顾2002年对2002~2010科学图谱发展的预测

下面三段预测是我在2002年在《Mapping Scientific Frontiers: The Quest for Knowledge Visualization》的结尾对未来10年的展望。现在回头看看哪些实现了，哪些还没有，还有哪些没有提到的。好象总体上乐观了点儿。。。

1。 2002~2005

实现自动引文索引，大规模文本分析，自动提取引文环境，自动生成多个引文环境的摘要，等等。

For thenext 3~5 years, between 2002 and 2005, several routes of research anddevelopment are likely to emerge or establish. In particular, the need forexpanding the current coverage of citation databases to include conferenceproceedings and patent databases can trigger further research in automaticcitation indexing and large-scale text analysis. The need for timelierdisciplinary snapshot should also drive much research onto this route. Automaticextraction of citation context will become increasingly popular. Softwareagents will begin to emerge for summarizing multiple citation contexts – animportant step in resolving the bottleneck of streamlining quantitative andqualitative approaches to science mapping. The recent surge of interest insmall-world networks is likely to continue. One can expect to see more specificstudies of scientific networks as small-world networks, including Web-basedresources analysis, traditional citation databases, and patent databases.Research in small-world networks is likely to draw much of attention to networkanalysis tools that can handle large-scale scientific networks. Cross-sectioncomparisons should increase.

2。2005~2010

面向知识的搜索和浏览工具，用户将用把更多的精力用在理解和解释信息而不是搜索信息，自动提取引文环境和摘要更为成熟，科学数据于科学文献相互连通

Between2005 and 2010, knowledge-oriented search tools and exploration tools willbecome widely available. Users major search tasks will probably switch fromdata-oriented search to comprehension and interpretation tasks. Intelligentsoftware agents will begin to mature for citation context extraction andsummarization.  Genomic maps will playmore substantial roles in linking scientific data and scientific literature. Asynergy of data mining in genomic map data and scientific literature willattract increasing interest.

3。2010~

科学预测和模拟开始应用到科学本身，用来辅助科学发现。

Beyond2010, mapping scientific frontiers should reach a point where science maps canstart to make forecasts and simulations. Powerful simulations will allowscientists to see the potential impact of a new technology. Further than that,we will have to wait and see.