It is believed that learning collocations is an effective way to enhance fluency. I wonder if there are any tools that, when provided with a text, can highlight its collocations? It would be great if it also can show degrees of words cohesion using different highlight colors.
Asked
Active
Viewed 314 times
3
-
@ChristopheStrobbe Thanks for your constructive contribution. – Real Dreams Jul 05 '17 at 22:03
-
If the purpose is to improve efficiency, it seems on-topic to me. I suppose that's based on the unspoken assumption that this practice actually does improve efficiency. If someone asked for a recipe search engine, ostensibly to improve fluency, I would consider that entirely off-topic. I'm curious to hear what others think. – Flimzy Jul 06 '17 at 12:13
-
1In my opinion, I think this is an on-topic resource request because such a tool could be used to learn any language. – fi12 Jul 07 '17 at 12:26
1 Answers
1
I am not aware of tools that highlight collocations in a given text, but there are a number of tools for collocation extraction that may be used to support this kind of functionality in other tools.
See for example:
- CoreNLP-jMWE: the "Stanford CoreNLP annotator implementing jMWE for detecting Multi-Word Expressions / collocations": this is a Java-based tool that uses a database generated from WordNet. (Its licence is GNU GPL 3, so there are some limitations on how you can use this in other projects.)
- mat kelcey's collocations, unfortunately not updated since 2001 and without a licence.
- Collocate: a commercial but inexpensive tool. (It is not clear when the site was last updated; it looks slightly neglected.)
- Collocation Extract (for Windows; free).
- Collocation Extract (by Wirote Aroonmanakun; for Windows; same tool as above).
- The paper Review of software applications for deriving collocations by Nikolaos K Anagnostou and George R S Weir reviews four software programs: WordSmith Tools 4 (WST 4), Collocate, Xaira and the Ngram Statistics Package (NSP)
- The paper Constructing a Collocation Learning System from the Wikipedia Corpus by Shaoqun Wu et al describes a collocation learning system named FlaxCLS that is part of the Flax system at the University of Waikoto, New Zealand. FlaxCLS is not a downloadable tool but a web interface that you can use to query the results of text mining work on Wikipedia at Flax. The same work also led to the development of a number of Android apps: FLAX Collocation Guessing, FLAX Collocation Matching and FLAX Completing Collocations.
With the tools that aren't open source, it is not clear how you would reuse their output in another program, except by exporting their output and then importing that output into a program that does the highlighting.
As a general conclusion, it appears that tools for identifying collocations require text mining on a sizeable text corpus and that such tools are mostly used by linguist (in the area of corpus linguistics). Although there are a few tools for the "general pubic", it is not clear whether they are still being maintained.
Update: Fraze finder is a tool to "hghlight collocations, chunks, idioms, phrasal verbs and semi-fixed lexical phrases in texts". The source code is available on GitHub.
Tsundoku
- 19,740
- 4
- 43
- 120
-
-
-
Yes it is. Yet does not answer the question that is looking for an already to use tool. – Real Dreams Jul 20 '17 at 02:00
-
@PHPst I have updated my answer with a link to a tool that may be useful: Fraze finder. – Tsundoku Sep 26 '17 at 16:07
-