Introduction Authentic Text Authentic Text ICALL (ATICALL) ICALL - PowerPoint PPT Presentation

ICALL: Part V ICALL: Part V Introduction Authentic Text Authentic Text ICALL (ATICALL) ICALL (ATICALL) Intelligent Computer-Assisted Language Learning Detmar Meurers Detmar Meurers Universit¨ at T¨ ubingen Universit¨ at T¨ ubingen ◮ The use of NLP in ICALL has primarily centered on Part V: Authentic Text ICALL (ATICALL) Introduction Introduction Pedagogical grounding Pedagogical grounding Exercise Generation & Information Retrieval for diagnosing learner errors and, more recently, testing The WERTi system The WERTi system and assessment. Language Learning Modeling FLT practice Modeling FLT practice Progression in WERTi Progression in WERTi Example 1: Pronouns Example 1: Pronouns Example 2: Passive ◮ Idea: Explore how NLP technology can support other Example 2: Passive Example 3: Adverb placement Example 3: Adverb placement Example 4: Tense and Aspect aspects of second language learning. Example 4: Tense and Aspect Realizing proposal Detmar Meurers Realizing proposal Creating ex. progression Creating ex. progression (Universit¨ at T¨ ubingen) Prototype Prototype ◮ Our specific focus: What can NLP contribute to Some challenges Some challenges Related approaches Related approaches awareness of language forms and rules, an important IR4LL (Ott 2009) IR4LL (Ott 2009) Measuring Text Difficulty component of adult second language acquisition? Measuring Text Difficulty based on joint research with Readability Formulas Readability Formulas Luiz Amaral, Vanessa Metcalf, Niels Ott Lexical Fequency Profiles ◮ WERTi: Automatic generation of language awareness Lexical Fequency Profiles Syntactic Complexity Syntactic Complexity (cf. Amaral, Metcalf, Meurers 2006; Metcalf, Meurers 2006, Ott 2009) Textbook structures activities based on real-world texts. Textbook structures A Search Engine Prototype A Search Engine Prototype ◮ IR4LL: Retrieval of authentic texts at the appropriate Information Retrieval Information Retrieval Adding Text Difficulty Adding Text Difficulty level for language learners Related Approaches Related Approaches European Summer School in Language, Logic, and Information Towards Evaluation Towards Evaluation Bordeaux. July 27–31, 2009 Summary Summary 1 / 54 2 / 54 ICALL: Part V ICALL: Part V Pedagogical grounding of our research Pedagogical grounding of our research Authentic Text Authentic Text ICALL (ATICALL) ICALL (ATICALL) Awareness The role of awareness Detmar Meurers Detmar Meurers Universit¨ at T¨ ubingen Universit¨ at T¨ ubingen Awareness (Schmidt 1995): Introduction Introduction ◮ Research on awareness shows: Pedagogical grounding Pedagogical grounding ◮ Noticing ◮ There is no learning without noticing. The WERTi system The WERTi system ◮ “conscious registration of an event” Modeling FLT practice Modeling FLT practice ◮ Awareness without input is not sufficient. Progression in WERTi Progression in WERTi ◮ low level of awareness Example 1: Pronouns Example 1: Pronouns ◮ “Learning takes place within the learner’s mind and Example 2: Passive Example 2: Passive ◮ implicit learning Example 3: Adverb placement Example 3: Adverb placement cannot be completely engineered by teachers or Example 4: Tense and Aspect Example 4: Tense and Aspect E.g.: noticing that sometimes speakers of Spanish omit the Realizing proposal syllabus designers.” Realizing proposal Creating ex. progression Creating ex. progression subject pronoun ◮ One can only provide opportunities for developing Prototype Prototype Some challenges Some challenges learner awareness. Related approaches Related approaches ◮ Understanding IR4LL (Ott 2009) IR4LL (Ott 2009) Measuring Text Difficulty ⇒ Consequences: Measuring Text Difficulty ◮ “recognition of a general principle, rule or pattern” Readability Formulas Readability Formulas Lexical Fequency Profiles ◮ Learners have to be exposed to linguistic features to Lexical Fequency Profiles ◮ higher level of awareness Syntactic Complexity Syntactic Complexity ◮ explicit learning Textbook structures acquire them. Textbook structures A Search Engine Prototype A Search Engine Prototype ◮ Learners have to notice those features. ◮ generalization can be internally generated or externally Information Retrieval Information Retrieval Adding Text Difficulty Adding Text Difficulty ◮ Tools presenting such linguistic features in a contextualized provided Related Approaches Related Approaches Towards Evaluation Towards Evaluation way, allowing for student interaction, can be helpful. E.g. understanding that Spanish is a pro-drop language Summary Summary 3 / 54 4 / 54