Archive

Archive for the ‘e-languages’ Category

Learning materials management: Links (1998-2004)

Originally implemented for a series of Canadian universities teaching Wirtschaftsdeutsch, then continually expanded into all of German for Queen’ s  University, and multiple languages, including non-western, for university of Michigan-Dearborn and Drake university.

Was based on an open source software project by Gossamer Threads popular for web 2.0 precursors of collaborative links collections, whose Perl-CGI code needed only minor modification to facilitate the “”commenting”” on instructor-“posted”  ( i.e.  assigned links) by students.

The model was Yahoo’s human-edited web-catalogue. the data structure was the tree (nested folders, unidirectional graph). For managing, I implemented a secondary branch mirroring the primary under the root “old links”  for, using Perl regex, automatically moving links which a batch link-checking management script in the open source had identified and logged as “broken” (404 and a few other similarly bad http return codes) into.

The original layout of the “ontology” first introduced me to the complexity of such a task. The basic content division was between 2 branches.

  1. web-based ready-made teaching materials for commenting (recommending, categorizing) by instructors and self-access by students (no feedback of student data to the instructor mostly, except by email, and outside of the application, in those days).
  2. the other content branch consisted of not teaching-related “”authentic materials””: the early day web applications, sometimes multimedia (maps, audio and video collections, news), often times also self-service database interfaces (online shopping and public services) whose language-wise rather restricted interface and topical focus (think Wirtschaftsdeutsch) lent themselves to capstone exercises at the end of textbook chapters (our “Friday in the lab””, not even a language lab then. Geek bonus points: one of these Fridays, a future queens university educated engineer asking me whether i had written all these pages they browsed through in the searchable catalogue of eventually 1500 links. Well, dynamic web pages were not common at all in education in those days, and the credit goes to Gossamer Threads.).

 

While there was hope to collect a comprehensive teaching resource through collaboration, “der Weg war das Ziel”, having students interact with and review foreign language web content. The links database  remained definitely, as it grow in bursts revolving around the topics of our chapters. I had a lot of fun finding instructional ways to having students review all those fancy web applications in which endless amounts of money were poured  before the first bubble in this millennium burst. E.g. the first early online city maps for “Wegbeschreibung”  in German 102. as well web 2.0 like developments, like grassroots web cams (Germans allowing the world to spy on their surroundings 24/7, including remote camera panning – you could go all kinds of places, “”Wie heißt der  bürgermeister von Wesel? was macht das wetter in der Schweizzzzzz?” but alas, the time lag, especially during winter term.

A couple of screen casts for instructor training are here and here.

Auralog Tell-me-more Demo Screencasts

Pedagogical rationale of timestretching audio for differentiating instruction

  1. Context: Higher Education in the UK has made considerable investments in digital lab infrastructure to improve second language instruction in times of deteriorating language take-up in the secondary sector, including widening participation. Digital language labs, apart from generic digital media, suffer from a lack of custom-made teaching materials that take advantage of the pedagogic features of the lab: grouping for personalization of teaching and learning. Pedagogical integration and development is needed to achieve the original intentions. A project to timestretch audio language learning materials for the digital audio lab promises integration software, pedagogical materials and, above all, a model of effective digital language lab use in teaching.
  2. Problem: In times of uneven language provision at the secondary school level and of shrinking language program sizes in HE, increasingly language teachers find themselves confronted with uneven language proficiency in their courses. Digital lab technology can help them to overcome the  “one size fits all” approach and personalize the students learning experience, for a greater inclusiveness in language programs and an increased proficiency boost for both the below and above average proficiency student groups.
  3. During my work with the language programmes at an English university, I could witness – and had to record – that the least proficient students, seeing themselves confronted with what was nowhere near “comprehensible input” (Krashen) for them, not only let the communication break down, but appeared so distressed that, despite being fully aware that their language output was being recorded as an assessment for the teacher to evaluate, started to curse and swear (in their native tongue) – while at the same time the upper portion of the class breezed through the exercise without any apparent difficulty.
  4. Proposed Solution:
    1. Technology to the rescue: The slowing down of digital audio – without pitch alteration –has been, while not a perfectly accurate representation of natural slow speech output, a popular benefit of digital technology in the language learning field for several years now (cf. e.g. Calico 2004), and I myself have experimented with it in the digital audio lab (Model imitation and Question – response exercises) and in publications (cf. Plagwitz, Karaoke in the Digital Audio Lab (2006)).
    2. What seems lacking are
      1. both an application that automates (by monitoring one of the network share directories that are part of the digital lab system) the slowing down (and speeding up) of audio for instructors (e.g. in 5% increments from 70% to 120% of original input) that are too time-pressed for producing materials, or even seeking out recordable on-air sources, and
      2. a model implementation in the digital audio lab (using dynamic grouping of students through the digital lab software) that creates exercises that would create exercises that can benefit from this approach (and can be shared), that applies them in a number of suitable (interpreting, ab initio language learning) modules and that assesses the proficiency improvement with this approach (using the outcome exam and a control group).
  5. Benefits: Greater fluency of both the least and most proficient students is to be expect after they were exposed to – as deemed fit by their instructors – slowed down/sped up exercises – ca. 20 exercises in the ab initio language learning module, practicing a small set of suitable new structures and vocabulary compared, with 2 control groups, and five interpreting rounds of 20-30 minutes. We will operationalize this by reusing regular assignment grading and use a control group, also of module-size, which must also use the digital audio lab, but with “one size fits all” audio.

A syllabus integrating language learning technologies into iLearning

iLearning means more intelligent

Producing and managing language learning content

produce-manage-content

Example 8: Auralog Tell-Me-More Speech Recognition Test

2008/08/29 1 comment

How usable is the Auralog Speech Recognition for language learning? This test, by a non-native speaker of English, gives some authentic data points.

image

The test shows: Auralog Speech Recognition

  1. can be easily tripped up; however, by errors that  a non-native language learner would not normally make
  2. more concerning is that the built-in AI, instead of e.g. escalating to additional feedback or help, like the pronunciation waveforms (which in itself seem to encourage only repeated attempts to mimic a given intonation, while not being fine-grained enough to spot mispronunciations on a word, let alone letter level) – lowers the requirements when a speaker repeatedly fails (which in extreme seems to amount to “waving through” any utterance).
  3. the preset dialogue – only few exercises including wrong answer options, most exercises testing only a comprehensible pronunciation of a given reading text which makes the exercise much easier for the built-in speech recognition, but also much less realistic and useful for a language learner (or more of a reading exercise).

Language Lab Techniques for (Self-)Evaluation and Grading of Student Recordings with Audacity

2008/08/18 2 comments

This quick and dirty (not narrated and uncut: time is money, and storage cheap…) video demonstrates a technique in (the free audio editor) Audacity with which instructors and students can more easily (self-)evaluate parallel recordings from (be it model imitation, question-response, or consecutive interpreting exercises in) the language lab (in this case the output of a Sanako Study1200, which automatically gets stored in a folder on network share):

When?

What?

0,00

how to load 10 student files à 5mb = 2:30min (but as a batch, allowing you do something else in the foreground instead of waiting)

2,50

how to select a part of the timeline to play

3,00

how to move tracks up to more easily work with them and the menu

3,30

how to play all tracks simultaneously (choir, normally not very useful for evaluation)

3,40

how to play only one track (solo): evaluate & compare

Automating Auralog Tell-Me-More with AutoIt. Presentation at EUROCALL 2008

Auralog Tell-Me-More is a leading language learning software system which provides a vast amount of content in an advanced technical infrastructure that we found lacking in usability within an higher education language learning environment.

AutoIt is a programming language for GUI automation which I used to better integrate the Auralog software into the higher education language learning process, including

  1. programmatic creation of courses and accounts
  2. programmatic extraction and digital repository management for over 30.000 learning units.Click to view a work sample from my portfolio
  3. programmatic creation of 10,000s of learning paths,

Results were presented (screencast) at EUROCALL 2008: “Automating Auralog (pdf)”:

    1. cpurse and account creation

creates 100s of courses , creates and enrols up to 2,000 student accounts every term,

  1. content extraction produces files for adding search and spreadsheet for sort/filter functionality:
  2. learning path creation.

More detailed background information here: plagwitz_auralog_accounts_project_pub.pdf, plagwitz_auralog_project_pub.pdf