AntWordProfiler = Paul Nation's Range program

I'm sorry that I've been too greedy to expect so many additional features. I know it is easier said than done. I really appreciate your endeavors in developing AntConc and AntWordProfiler and, most importantly, made them all free.

One last bit of my greed is the implementation of concgramming in AntConc. Or alternatively, you can make a standalone tool for concgramming, because concgramming is computationally expensive. It might be a good idea to follow WS to concgram on a previously indexed file.
 
I can add a stop lists function similar to AntConc quite easily.

The mark texts function is useful. Perhaps I can implement this in a modern way, with standard tags, e.g.

This_1 is_1 a_1 difficult_2 thing_1 to_1 implement_3

This allows the resulting file to be used in other software, e.g AntConc.

I haven't looked at how to implement ConcGrams but I don't think it is difficult. I agree that a separate program would probably be better.
 
Re: Re: AntWordProfiler = Paul Nation's Range program

I can add a stop lists function similar to AntConc quite easily.

The mark texts function is useful. Perhaps I can implement this in a modern way, with standard tags, e.g.

This_1 is_1 a_1 difficult_2 thing_1 to_1 implement_3

--I used to read Range-marked texts into AntConc to do a concordancing in order to build an off-the-list word list, because that's useful for textbook compilation when we wish to add a new word list after each text.

--Can you enable list export to save the words while highlighting them in different colors? e.g. which words fall into the 1st 1000 words, and/or which are off the list in the texts?

--Does AntWordProfiler support phrase list search, because Range only supports word matching?
e.g. a dummy phrase list
a lot of
at the end of
on top of
more often than not
of course

--Does AntWordProfiler support regex lists? In other words, can regular expressions be used as base word lists as we do with file-based concordancing? In that case, complex grammatical patterns can be profiled. AWP will definitely get an upper hand over Range if phrase list and regex list matching can be implemented.
e.g. a dummy regex list
\w+_N\w+\s\w+_V\w+
\w+_VB\w+\s+\w+_V\wN
\w+_I\w+\s\w+_N\w+\w+_I\w+

This allows the resulting file to be used in other software, e.g AntConc.

I haven't looked at how to implement ConcGrams but I don't think it is difficult. I agree that a separate program would probably be better.

--Currently, Chris Greaves and Mike Scott have two ConcGram tools, and yours will be the first free concgramming tool ever.
 

 --Can you enable list export to save the words while highlighting them in different colors? e.g. which words fall into the 1st 1000 words, and/or which are off the list in the texts?

I'm not sure what you mean here. Do you want a file with just the words and not the original target file?

Maybe you could show me the output that you want.
--Does AntWordProfiler support phrase list search, because Range only support word matching?

Good question. I'm not sure! If it doesn't work with lists now, it would be very easy to do. Maybe just one or two lines changed in the code.

--Does AntWordProfiler support regex lists? In other words, can regular expressions be used as base word lists as we do with file-based concordancing? In that case, complex grammatical patterns can be profiled.

Good question. Again, one or two lines changed would allow this to work. The main problem is how to document it so that novice users know what is happening. More features are not always a good idea.


--Currently, Chris Greaves and Mike Scott have two ConcGram tools, yours will be the first freeware ever.

Great. Let me work on this.
 
Re: Re: AntWordProfiler = Paul Nation's Range program

--Can you enable list export to save the words while highlighting them in different colors? e.g. which words fall into the 1st 1000 words, and/or which are off the list in the texts?

I'm not sure what you mean here. Do you want a file with just the words and not the original target file?

Maybe you could show me the output that you want.

For example, I want all red words to be saved into a freq. list file, all green words to be a freq. list file, all blue words to a freq. list file, and non-level words to a freq. list file.
The green words are high-freq., likely easy, words which deserve less attention; non-level words are new words in the current text which are worth special explanation or practice.

This is most useful to EFL/ESL textbook compilation and for test and exercise generation purposes.
 

附件

  • result_in_color.jpg
    result_in_color.jpg
    124.4 KB · 浏览: 4
回复: Re: AntWordProfiler = Paul Nation's Range program

--Does AntWordProfiler support regex lists? In other words, can regular expressions be used as base word lists as we do with file-based concordancing? In that case, complex grammatical patterns can be profiled.

Good question. Again, one or two lines changed would allow this to work. The main problem is how to document it so that novice users know what is happening. More features are not always a good idea.

I can help with your documentation if I'm trusted at all.
 
For example, I want all red words to be saved into a freq. list file, all green words to be a freq. list file, all blue words to a freq. list file, and non-level words to a freq. list file.
The green words are high-freq., likely easy, words which deserve less attention; non-level words are new words in the current text which are worth special explanation or practice.

You can already do this in the standard display, but you have to only open one file at once. (Batch processing would be better). Below, are a sample of the "NOT in lists" results for a single text. The 1000 level, 2000 level etc. appear above this.

Groups NOT Found In Base Lists

Group Range Freq uf_1
metropolitan 1 1 1
naples 1 1 1
paris 1 1 1
pompeii 1 2 2
romans 1 2 2
rome 1 1 1
seashore 1 1 1
york 1 1 1
 
Last edited:
回复: AntWordProfiler = Paul Nation's Range program

请问： 如果用AntWordProfiler分析一个教材的词汇分布，用程序中提供的3个level lists就够了吗？还是不是需要更多的lists呢？哪里有现成的level lists?
 
You must log in or register to reply here.
Back
顶部