For English there are quite a few reasonably well-written applications. With German and French I'm rather lost. I haven't encountered any monolingual extractors for these languages, yet. And I haven't found any reliable + affordable bilingual term extractors, either. As a conference interpreter I'd love to extract difficult terms from (parallel) texts in the nick of time.
I'm pretty sure there are some software kits to found that do term extraction, but if you want to get the best possible results, you should think about implementing your own term extractor. I could give you some papers on that, using different linguistic methods plus methods such as tf*idf or maybe the loglikelihood measure.
Das würde mich SEHR interessieren. Ich bin zwar kein software engineer aber ich habe früher viele Anwendungen (in VisualBasic 6.0) geschrieben, die ich für die Analyse und Bearbeitung eigener Dokumente einsetzte. Wenn die von Dir vorgeschlagenen Methoden mich nicht überfordern, würde ich sie gern ausprobieren, oder zumindest das ihnen zugrundeliegende Konzept des näheren anschauen. Danke im voraus für die papers und Vorschläge.
Termextraktion:
Antti Arppe (1995): Term Extraction from Unrestricted Text. In: Proceedings of NODALIDA-95. [http://www2.lingsoft.fi/doc/nptool/term-extraction.html].
John S. Justeson, Slava M. Katz (1994): Technical Terminology: Some Linguistic Properties and an Algorithm for Identification in Text. In: Natural Language Engineering 1(1), S. 9-27.
Ulrich Heid (2000): A Linguistic Bootstrapping Approach to the Extraction of Term Candidates from German Text. In: Terminology, S. 161-181.
Heike Bieler, Stefanie Dipper (2008): Measures for Term and Sentence Relevances: an Evaluation for German. In: Proceedings of LREC 2008.
Hans-Friedrich Witschel (2005): Text, Wörter, Morpheme – Möglichkeiten einer automatischen Terminologie-Extraktion. In: Proceedings of GLDV 2005.
TF*IDF:
http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.97.7340
Loglikelihood:
tba
And especially for French and German: http://www.ims.uni-stuttgart.de/~gojunaa/publ/variation_gscl_2011.pdf
Vielen Dank auch von mir für die interessanten Literaturhinweise. Ich beschäftige mich mit diesem Thema in meiner Masterarbeit. Justeson/Katz und Heid werden praktisch in allen Quellen zum Thema Terminologie-Extraktion zitiert. Diese Aufsätze muss man gelesen haben. Witschel war mir auch ein Begriff ("Terminologie-Extraktion: Möglichkeiten der Kombination statistischer und musterbasierter Verfahren"), aber ich bin auf den von Ihnen empfohlenen Aufsatz sehr neugierig.
We have developed a method called Likey (Language-Independent
KEYphrase extraction) based on the use of reference corpora. Likey has a very light-weight preprocessing and no language-specific resources are needed in addition to the reference corpus. Thus, the method is not restricted to any single language or language family. We have been very pleased with the results that have been based on experiments with more than ten languages and used the method in various applications. The method is presented in a Coling paper that you can find at
http://users.ics.aalto.fi/tho/info/paukkeri2008coling.shtml
The method and its use is discussed in some detail in Mari-Sanna Paukkeri's recent PhD thesis (Section 4.1):
http://lib.tkk.fi/Diss/2012/isbn9789526048345/isbn9789526048345.pdf
We have applied the Likey method, for instance, in assessing user-specific difficulty of documents:
http://www.sciencedirect.com/science/article/pii/S0306457312000532
Termostat (http://olst.ling.umontreal.ca/~drouinp/termostat_web/interfaceTermostat.php) performs monolingual term extraction in French (as well as in English, Spanish, Portuguese and Italian). You need to create a user account, but it is fast and free.
Dear Robert,
You are welcome to test our Beta terminology services https://demo.taas-project.eu!
We opened this for Beta testing last week and are looking forward to our users' feedback. The Beta is free of charge, you can create your project, add your documents, customise extraction (language specific and language independent extraction are both available), get source term candidates and their translation candidates from various sources, refine extraction results, and share your project with other users.
By default your project is private and is visible only for you, but you might also want it to publish it so that other users could see your terminology.
Now the service supports 8 languages, and FR and DE are among them. During next months we will add 17 more languages covering EU official languages and RU.
Enjoy the TaaS services and do not hesitate to ask me questions!
Yours, Tatiana