Archive

Archive for the ‘service-is-evaluating-learning-tools’ Category

How to download Centro Spanish Textbook audio

The audio is in (compressed) mp3 format. Just right-click on any audio link and choose save target/link as (or similar, depending on browser), like so: clip_image001

Example from Centro – Puntos de partida: Online Laboratory Manual, 8th Edition (you need access privileges to follow this link, but you can send links around, users can open them, provided their webbrowser is already logged into the centro site  – getting access and finding you way around the website is the real issue. Webspiders are prohibited, though – and even Downthemall saves only files without extensions: rename them to .mp3 or save them manually as .mp3 in the first place, as shown above. Does not work as above? Try a different web browser.

Query treebanks with Fangorn for English SLA?

To provide inductive empirical examples,  SLA  classes have benefitted from query interfaces to target language text corpora in SLA. But corpora are usually POS-tagged – and queried – at best, which constitutes a certain “impedance mismatch” to what SLA classes actually teach. The Fangorn very large treebank query language beta demonstration page

image

looks already interesting for analyzing English in SLA (hover over tree elements to highlight the corresponding text), including, thanks to its capability of editing and refining queries graphically from the search results, for demonstrations during face-to-face classes. Wondering whether other corpora than Penn Treebank, Wikipedia (5k and 5000k sentences) will be made available online, and other languages but English will be supported.

Example 7: Exercise dictating in German to an LRC Windows 7 computer

How can we get language students more speaking practice with qualified, but affordable feedback ? Native speaker contact remains difficult to organize even in the days of online conferencing. The LRC hosts language tutoring, but numbers are limited. Enter speech recognition, the holy grail of iCALL,  much easier for learners to relate to than the voice graph that digital audio can be broken down to, and thus for a long time a standout feature of costly second-language-acquisition packages like Auralog Tell-me-More (speech recognition in English tested here) – but now the LRC has Windows 7 Enterprise (and its free add-on language packs), and another crucial prerequisite: headphones with excellent microphones.

We are setting up the new Windows 7 computers in the LRC to allow for speech recognition in Chinese, English, French, German, Japanese and Spanish. Here is an example of me using this facility for a practicing my German during a dictation exercise:

Granted, German is my native tongue; but the example text is from the online component for the final chapter  of the “Treffpunkt Deutsch” 1st-year textbook in use here, which sends the readers to the website of the Swiss (-German) employment agency.

Apart from infrequent words ("Archiven") and Lehnwörtern ("Bachelor" etc.), Windows 7 speech recognition accuracy seems quite impressive. The above example was actually my first dictation, except that immediately beforehand, I invested a few minutes into the standard Windows 7 speech recognition training (aimed at training the user, although may behind the scenes teach the computer a few things about the speaker already also) and a few more minutes of voice training (this one is meant exclusively for the computer, but the user can also see it fail and why). The – rather simple trick to boost speech recognition results – certainly accessible to our students – seems to be to speak not only  clearly, but also slowly, with short pauses between most words.

Speech recognition in these languages is a feature of the Windows 7 (Enterprise/Ultimate version) “language packs” that we installed and switched to – that is why the entire computer interface appears in German. Practicing the L2 with (computer—operating) “voice commands” (instead of with a mouse) is also possible, simpler than replacing the keyboard (mostly) by voice, but not as easy to devise homework exercises for.

Tips for designing exercises using speech recognition: As the example shows ("Archiven") , doing all corrections by voice can quickly become tedious. But there is no pedagogical need to have your students’ bang their heads against this wall. Instead, just ask your students to correct their automatically recognized words manually at the end of their video, after their dictation. This way both you and your students get a clear summary of what they achieved – even clearer if they dictate in MS-Word with the spell and grammar check for the language (automatic with the switch to the language pack for the language) and (using key combination CTRL+SHIFT+E) track changes. We will show you later TBA:how we now enable students to easily record their screen and TBA:upload their screencast into Moodle Kaltura.

Part-of-speech tagging

A few notes on POS-tagsets, collected online. You can view a larger version here.

Comparison of NLP Platforms

Not really a comparison, only a notebook compiled from online sources. Not really fit for publication either, unless “sharing is caring”. You can view a larger version here.

How to stream video clips to students in classroom and at home, using Moodle Kaltura

  1. DVDs are getting a bit long in the tooth, not to mention VHS, and can form a real obstacle or time-consuming distraction in an educational setting, from handling the media to finding compatible software and/or hardware players for the media.
  2. Fortunately, there is a now a better way to make video clips available to students than uploading them to YouTube.com:
    1. university-supported,
    2. more compliant with copyright and fair use restrictions (which still apply)
    3. also requiring only a web browser (available on all campus computers, including teacher computers in classrooms, including those that have no (region-free) DVD-player installed)
    4. and a course enrolment. But access to a Moodle course can  now be considered a given, both for teachers and students.
  3. Moodle Kaltura allows for easy
    1. uploading of a video file by the teacher
    2. viewing by the student (streamed – Flash required, not different from YouTube.com).
  4. View a screencast example how easy it is with Moodle Kaltura to upload and playback a video clip from a movie DVD.
    1. Not different from YouTube.com, you still need to edit out the segment from the DVD that you want to show in your class,  uploading a full DVD I do not intend to test.
    2. From this example, you can also get an idea how long the server-side encode takes before the video an be streamed back to students: the short clip of a few minutes here starts playing back at 12:40. Naturally, a teacher would prepare their course, including all video uploads, before the term starts or possibly before the week starts, or, in extremis, before the class starts – in practice, only the – extremely unlikely – scenario where the teacher would try and upload the video during the class is not supported.

Protected: Spring 2012 Faculty Workshop I: How to ease your end-of-term oral assessment burden with the help of the LRC Moodle Kaltura and Sanako Study 1200 oral assessments

2012/04/06 Enter your password to view comments.

This content is password-protected. To view it, please enter the password below.

Linguee dictionary lookup based on parallel corpora

2012/04/04 2 comments
  1. Support for more languages is planned (Chinese, Japanese):linguee parallel corpora dictionary lookup
  2. linguee parallel corpora dictionary lookup1
  3. The interesting approach based on parallel corpora provides a wealth of empirical data, albeit a bit raw  and of varying quality, e.g.:
  4. linguee parallel corpora dictionary lookup griff in die wundertüte