IIS: Rapid Development of a Frame Semantic Lexicon

国际计算机科学研究所的框架项目正在构建一个独特详细的英语词典，既可供人类阅读，又可供计算机用于自然语言处理（NLP）。它基于由charles Fillmore提出的框架语义学理论，该理论认为，许多单词的含义最好通过定义事件或状态中参与者和道具集合的语义框架来理解。例如，在“Alicia煮土豆20分钟”中，“煮”这个词让人联想到Alicia是厨师，土豆是食物，20分钟是持续时间——但也有某种容器（锅？）和某种加热工具（炉子？）这些没有被提及，但却是框架的重要组成部分。框架的范围从非常简单的（如放置）到非常复杂的（如复仇）。自1997年以来，FrameNet项目一直在描述语义框架（目前为750个，包含大约10,000个词义），并对唤起它们的句子（135,000个）进行注释，显示哪些单词在框架中扮演了哪些角色。然后，注释的句子被自动制表，以产生一个词汇条目，举例说明每个引起框架的单词的所有可能的句法模式。该词典被用于世界各地数百个NLP研究项目；其他语言的类似项目也在进行中。然而，这种方法非常耗费人力，需要仔细检查语料库中的句子，以保证提取和注释所有的语义和句法组合。这个奖项是为了表彰一个为期三年的努力，该努力将用于分组语料库示例的复杂工具（由Adam Kilgarriff开发）集成到框架提取和注释工作中，使工作人员能够更快地找到语法模式并半自动地注释示例，从而迅速增加所涵盖的单词数量，并使框架词典在更实际的NLP应用中有用。欲了解更多信息，请访问http://framenet.icsi.berkeley.edu。

The FrameNet Project at the International Computer Science Instituteis building a uniquely detailed lexicon of English, both readable byhumans and usable by computers for natural language processing(NLP). It is based on the theory of Frame Semantics, originated byCharles Fillmore, which recognizes that the meanings of many words arebest understood in relation to semantic frames defining the setof participants and props in an event or state. For example, in"Alicia boiled the potatoes for 20 minutes", the word "boiled" evokes aframe involving Alicia as the Cook, the potatoes as the Food, and 20minutes as the Duration---but also some kind of Container (a pot?) andsome Heating instrument (a stove?)--these are not mentioned, but areessential parts of the frame. Frames range from very simple, likePlacing, to very complex, like Revenge.Since 1997, the FrameNet project has been describing semantic frames(currently 750, containing roughly 10,000 word senses), andannotating sentences which evoke them ( 135,000), showing which wordsplay which roles in the frame. Annotated sentences are thenautomatically tabulated to produce a lexical entry exemplifyingall possible syntactic patterns for each frame-evoking word. Thislexicon is used in hundreds of NLP research projects around the world;and similar projects are underway for other languages.This methodology is, however, very labor-intensive, requiring carefulexamination of sentences in the corpus to guarantee that all semanticand syntactic combinations are extracted and annotated. This award isfor a three-year effort to integrate sophisticated tools for groupingcorpus examples, (developed by Adam Kilgarriff), into the FrameNetextraction and annotation work, enabling staff to find syntacticpatterns faster and annotate examples semi-automatically, thus rapidlyincreasing the number of words covered and making the FrameNet lexiconuseful in more practical NLP applications. For further information,visit http://framenet.icsi.berkeley.edu.