TUTORIAL DIALOG MD. FAISAL MAHBUB CHOWDHURY LECTURER: DR. IVANA KRUIJFF-KORBAYOVÁ Advanced Dialogue Modeling For Practical Application’08-09 Tutoring MD. Faisal Mahbub Chowdhury Tutorial Dialog 8 th January, 2009 Discussion or demonstration Learner’s knowledge assessment Construction and evaluation of knowledge at the same time Socratic vs. didactic ! Didactic teaching involves spoken instruction. Teachers are central to this method of teaching because they control the content and layout of the entire session. Learning is generally passive, by watching and listening during the lecture, but students also have the opportunity to make notes for use in future learning. [Gray 2007] ! Socratic method is a form of philosophical inquiry in which the questioner explores the implications of others' positions, to stimulate rational thinking and illuminate ideas. [Wikipedia - http://en.wikipedia.org/wiki/Socratic_method] MD. Faisal Mahbub Chowdhury Tutorial Dialog 8 th January, 2009 Modes of tutoring/ Types of tutoring services ! One-on-one tutoring ! Supplemental Instruction ! In-Class Tutoring ! Workshops ! Mentoring ! Referrals ........ ...... .... Details are available here - http://www.castutoring.neu.edu/tutoring_services/types_tutoring/ MD. Faisal Mahbub Chowdhury Tutorial Dialog 8 th January, 2009
13
Embed
TUTORIAL DIALOG M od esfturing/Typ CHOWDHURY … · TUTORIAL DIALOG MD. FAISAL MAHBUB CHOWDHURY LECTURER: DR. IVANA KRUIJFF-KORBAYOVç Advanced Dialogue Modeling …
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
TUTORIAL DIALOGMD. FAISAL MAHBUBCHOWDHURY
LECTURER:DR. IVANA KRUIJFF-KORBAYOVÁAdvanced Dialogue Modeling For PracticalApplication’08-09
! Didactic teaching involves spoken instruction. Teachersare central to this method of teaching because theycontrol the content and layout of the entire session.Learning is generally passive, by watching and listeningduring the lecture, but students also have theopportunity to make notes for use in future learning. [Gray2007]
! Socratic method is a form of philosophical inquiry inwhich the questioner explores the implications of others'positions, to stimulate rational thinking and illuminateideas. [Wikipedia - http://en.wikipedia.org/wiki/Socratic_method]
BEETLE (Basic Electricity and Electronics Tutorial Learning
Environment)
! The tutoring tactics are designed based on the analysis of a collection of dialogueswhere the students performed labs using GUI with the help of human tutors. Tutorsand students communicated from different machine through chat. Tutor couldmonitor student's progress from her machine.
! The teaching tactics are identified and annotated in these dialogues for training thesystem.
! Some of the human teaching tactics identified from the data are –! If a student rephrases a wrong answer for a question again, the tutor does not repeat the
same question and tries a different utterance.
! Tutor is careful not to perform steps of a teaching tactic that are unnecessary.
! Sometimes, even if the tutor thinks the student knows the instructions, the tutor does thefollowings –
! If the student types input then the parser produces anunderspecified logical form that the interpreterattempts to fully specify.
! The interpreter uses the problem solving manager,dialogue context, expectations, and curriculum toevaluate input (note, input to the tutor may simply bethat the student is idle).
! The interpreter updates the student model.
! The interpreter may send messages directly to thedialogue planner (e.g., an evaluation of the student’sanswer to a question or an alert when one of thevalues in the student model falls below threshold)
! If a conversational game is in progress, then theconversational game engine runs it, else the reactiveplanner is run to load a new conversational game.
! The reactive planner is run if a conversational gameends or there is unexpected student input (e.g., thestudent says red wire instead of red lead).
! Authors have videotaped over 100 h of naturalistic tutoring,transcribed the data, classified the speech act utterances intodiscourse categories, and analyzed the rate of particulardiscourse patterns.
! These analyses revealed that human tutors rarely implementintelligent pedagogical techniques such as bona fide Socratictutoring strategies, modeling–scaffolding–fading, reciprocalteaching, frontier learning, building on prerequisites, cascadelearning, and diagnosis/remediation of deep misconceptions.
! Instead, tutors tend to coach students in constructingexplanations according to the EMT dialogue patterns.
! The EMT dialogue strategy is substantially easier to implementcomputationally than are sophisticated tutoring strategies.
! Although, an ideal answer (for the type questionAutoTutor asks a student) is approximately three toseven sentences in length, the initial answers to thesequestions by learners are typically only one word totwo sentences in length.
! AutoTutor engages the learner in a dialogue thatassists the learner in the evolution of an improvedanswer that draws out more of the learner’s knowledgethat is relevant to the answer.
! The dialogue between AutoTutor and the learnertypically lasts 30–100 turns.
The underlying idea of LSA is that the aggregate of allthe word contexts (inside a given large corpus) inwhich a given word does and does not appearprovides a set of mutual constraints that largelydetermines the similarity of meaning of words and setsof words to each other [Landauer et al. 1998].
In other words, LSA is a statistical technique thatmeasures the conceptual similarity of any two piecesof text, such as words, paragraphs, sentences etc.
! Collagen is a middlewaresystem based on collaborativediscourse.
! Collagen maintains a model ofthe discourse state shared bythe user and the computeragent.
! Agents constructed usingCollagen use the discoursestate to generate an agenda ofcandidate discourse acts,including both utterances anddomain actions, and thenchoose one (discourse act) toutter or perform.
! Collagen’s declarative languagecan be used to representdomain-specific proceduralknowledge. This knowledgeserves as a model of how domaintasks should be performed.
! Each task is associated with oneor more recipes (i.e., proceduresfor performing the task).
! Each recipe consists of severalelements drawn from a relativelystandard plan representation.
! Collagen model of collaborative dialogue includes two mainparts:! a representation of discourse state and
! a discourse interpretation algorithm, which specifies how to updatethe discourse state given a new action or utterance by either the useror agent. Its objective is to determine how the current act contributesto the collaboration.
! Collagen partitions the discourse state into three interrelatedcomponents:! linguistic structure,
! The linguistic structure groups the dialogue history into ahierarchy of discourse segments. Each segment is acontiguous sequence of actions and utterances that contribute tosome purpose (e.g., performing a subtask).
! The attentional state represents what the user and agentare talking about or working on at a given moment. It isrepresented as a stack of discourse purposes, called the focusstack. When a new discourse segment is begun, its purpose ispushed onto the stack. When a discourse segment is completed ordiscontinued, its purpose is popped off the stack.
! The intentional structure represents the decisions thathave been made as a result of the actions and utterances(reflected by the linguistic structure and attentional state),independent of their order. The intentional structure isrepresented as plan trees. Nodes in the tree represent mutuallyagreed upon intentions (e.g., to perform a task), and the treestructure represents the sub-goal relationships among theseintentions.