APTct

APTct compares phonetic transcriptions using International Phonetic Alphabet (IPA) symbols to provide quantitative, phonologically informed alignment and scoring for linguistic and clinical research.


Key Features:

  • IPA and extIPA symbol support: Accepts the full range of International Phonetic Alphabet (IPA) symbols and most extIPA symbols as input for comparison.
  • Modified edit distance algorithm: Uses a modified edit distance algorithm to compute quantitative distances between transcriptions.
  • Phonological alignment: Applies phonological alignment principles to inform the alignment step prior to comparison.
  • Alignment and scoring visualizations: Produces visual representations of optimal alignments and the scoring operations performed by the algorithm.
  • Validation against expert scoring: Has been tested for agreement with expert hand scoring, with minor discrepancies attributed primarily to input errors.

Scientific Applications:

  • Clinical linguistics: Facilitates analysis of speech disorders by enabling quantitative comparison between patient transcriptions and expected pronunciations.
  • Phonetic research: Enables detailed analysis of phonetic variation across languages and dialects using IPA/extIPA transcriptions.
  • Educational assessment: Supports evaluation and comparison of student transcriptions in phonetics training and assessment contexts.

Methodology:

APTct applies a modified edit distance algorithm informed by phonological alignment principles to align and compare IPA and extIPA transcriptions and generates visualizations of optimal alignments and scoring operations; validity was assessed by comparison to expert hand scoring.

Topics

Details

Tool Type:
web application
Added:
6/14/2021
Last Updated:
8/13/2021

Operations

Publications

Bailey DJ, Speights Atkins M, Mishra I, Li S, Luan Y, Seals C. An automated tool for comparing phonetic transcriptions. Clinical Linguistics & Phonetics. 2021;36(6):495-514. doi:10.1080/02699206.2021.1896783. PMID:33715568.

PMID: 33715568
Funding: - Auburn University: College of Computer Science and Software Engineeri, College of Liberal Arts Stevens Research Fund, Undergraduate Research Fellowship Award from the O - National Science Foundation: Grant CNS-1457855 / UFDSP00010405 to Seals