Assess language proficiency Workflow Blueprint

Choose a representative text, audio, or video sample that matches the target proficiency level and context (e.g., academic, professional, conversational). Ensure the sample is authentic and unaltered to reflect real-world language use. Pre-process the material by cleaning any noise, transcribing audio if needed, and confirming content authenticity.

Why Google Cloud Speech-to-Text: Google Cloud Speech-to-Text provides real-time streaming transcription and batch audio file processing with speaker diarization, ideal for converting authentic spoken language samples into text for analysis.

Establish clear assessment criteria based on a recognized framework (e.g., CEFR, ACTFL, or custom rubric). Break down proficiency into dimensions such as grammar, vocabulary, fluency, coherence, and pronunciation. Create a scoring rubric with descriptors for each level (e.g., A1–C2) to ensure consistent evaluation.

Why Notion AI 3.0: Notion AI 3.0 can build custom AI agents to search across frameworks and generate structured rubric templates, while also providing a workspace to organize proficiency criteria.

Use annotation tools to mark specific linguistic features in the sample, such as errors, complex structures, or vocabulary usage. For written text, highlight grammatical issues and lexical choices; for spoken samples, note pronunciation and fluency markers. Record observations in a structured annotation layer (e.g., comments, tags, or a spreadsheet).

Why Prodigy: Prodigy enables Named Entity Recognition and Text Classification, allowing systematic annotation of linguistic features (e.g., parts of speech, errors) in the language sample.

Compare the annotated observations to the rubric descriptors and assign a level or score for each dimension. Be systematic: for each dimension, find the descriptor that best matches the sample's performance. Record scores in a structured format (e.g., table or scoring sheet) and note any borderline cases.

Why Arcwise AI: Arcwise AI enables natural language formula generation and automated data cleaning in spreadsheets, ideal for creating a scoring sheet and calculating dimension scores against the rubric.

If the assessment involves comparing proficiency across languages or checking translation accuracy, produce a faithful translation of the sample. Use a professional translator or AI tool with human review to ensure the translation captures nuances. This step is optional and only needed when the assessment requires bilingual analysis.

Why Microsoft Translator: Microsoft Translator supports text, speech, and document translation across multiple languages, enabling cross-linguistic comparison of the language sample.

Compile the scores, annotations, and justifications into a clear proficiency report. Then, create an interactive quiz (e.g., multiple-choice, fill-in-the-blank) based on the sample to test the learner's own understanding or to reinforce assessment findings. Use quiz-building tools to embed the sample and questions, and include feedback for each answer.

Why ProProfs Quiz Maker: ProProfs Quiz Maker allows creation of scored quizzes with AI-generated questions, ideal for building an interactive proficiency assessment based on the report.