请问wsmith4计算MI值只需要INDEXER TOKEN 即可吗
- 主题发起人 dana922
- 时间
回复: 请问wsmith4计算MI值只需要INDEXER TOKEN 即可吗
You can compute MI scores using the indexed wordlist. Collocations can of course include function words. And among the few contents words is "acciddent". Without the full corpus at hand, if you want to know what is "happening", you can use the online FLOB concordancer - after excluding the instances of "happen* to" which means "by chance" - you will find what is happening typically refers to accidents, something aweful,somethin nasty, and "the inevitable" (If something is desired, why should one try to avoid it?).
You can compute MI scores using the indexed wordlist. Collocations can of course include function words. And among the few contents words is "acciddent". Without the full corpus at hand, if you want to know what is "happening", you can use the online FLOB concordancer - after excluding the instances of "happen* to" which means "by chance" - you will find what is happening typically refers to accidents, something aweful,somethin nasty, and "the inevitable" (If something is desired, why should one try to avoid it?).
回复: 请问wsmith4计算MI值只需要INDEXER TOKEN 即可吗
to xu,
MI大于3表示两词是显著搭配词, 现学现卖了
to xiao,
我首先在wordlist中打开FLOB index, do lemma matches, and then compute relationships, 但显示的依然是happen, happened, happening, happens, 各自的MI，不是HAPPEN的，里面没有accident, 甚至没有content words,直觉上感觉不对，但不知哪里出了问题。
CLEC中HAPPEN出现400余次，FLOB中只有300余次。
在on FLOB concordance搜到的检索行没有太多context, 但我已经很满意了，呵呵。
to xu,
MI大于3表示两词是显著搭配词, 现学现卖了
to xiao,
我首先在wordlist中打开FLOB index, do lemma matches, and then compute relationships, 但显示的依然是happen, happened, happening, happens, 各自的MI，不是HAPPEN的，里面没有accident, 甚至没有content words,直觉上感觉不对，但不知哪里出了问题。
CLEC中HAPPEN出现400余次，FLOB中只有300余次。
在on FLOB concordance搜到的检索行没有太多context, 但我已经很满意了，呵呵。
Last edited:
