XSLTを用いた教室談話構造可視化のためのXMLスキーマの設計

全文

(1)Title. XSLTを用いた教室談話構造可視化のためのXMLスキーマの設計. Author(s). 片桐, 徳昭; 河合, 剛. Citation. 北海道教育大学紀要. 人文科学・社会科学編, 66(2): 1-16. Issue Date. 2016-02. URL. http://s-ir.sap.hokkyodai.ac.jp/dspace/handle/123456789/7887. Rights. Hokkaido University of Education.

(2) 北海道教育大学紀要（人文科学・社会科学編）第66巻第２号 Journal of Hokkaido University of Education（Humanities and Social Sciences）Vol. 66, No.2. 平成 28 年２月 February, 2016. Designing an XML schema for classroom discourse visual representation through XSLT KATAGIRI Noriaki and KAWAI Goh＊ Department of English, Asahikawa Campus, Hokkaido University of Education, English Communication Studies ＊. Research Faculty of Media and Communication, Hokkaido University. XSLTを用いた教室談話構造可視化のためのXMLスキーマの設計片桐徳昭・河合剛＊北海道教育大学旭川校英語コミュニケーション学研究室＊. 北海道大学大学大学院メディア・コミュニケーション研究院. ABSTRACT Building upon the core notion of the classroom discourse hierarchical structure proposed by Sinclair and Coulthard (1975), we designed a set of an extensible mark-up language (XML) schema and an XSL transformation (XSLT) style sheet for annotating speech occurring in classrooms for learning non-native languages. The XML schema categorizes and groups the function of utterances (e.g., instruction, query, response) for quantitative and qualitative analyses of exchanges between teachers and students. The structure and content of classroom discourse are expressed in XML, which through an XSLT style sheet is converted into a table suited for discourse analysis. We have succeeded in designing a classroom discourse hierarchical structure in an XML schema in which classroom teacher-student interactions are aligned with discourse attributes. We yielded visual classroom discourse structure transcripts using our schema through XSLT. We hope that our annotation method and the visual representation of the classroom discourse will allow teachers to discover, for instance, that although teachers encourage students to be more vocal, in fact teachers do most of the talking; or that heretofore unnoticed inter-student conversations benefit learning.. 1. Introduction. skills by analyzing their classroom discourse elements that teachers are unlikely to observe. 1.1 Classroom corpora. during class. Annotated classroom corpora may. Classroom corpora help teachers develop their. be subjected to discourse or conversation. 1.

(3) KATAGIRI Noriaki and KAWAI Goh. analyses, or socio-cultural theories (O’Keeffe et. classroom discourse. Francis and Hunston. al, 2007).. (1992) raised our awareness of intonational. Classroom corpora require transcriptions of. phonology, particularly noting that acts always. classroom recordings, mostly consisting of. begin with a new tone unit (but see Brazil. spontaneous speech spoken by multiple. (1994, 2001), who points out that tone units. individuals, with frequent speech overlaps. require pauses, but pauses are not consistently. between speakers. Hence, corpora sizes are. found in classroom speech). Theoretical. typically much smaller compared to large-scale. validation of the IRF model includes Brazil. written language corpora. For instance, one. (1995), who explained the IRF model’s logical. classroom corpus contains approximately. process. Malouf (2001) developed the IRF model. 120,000 spoken words (Katagiri & Kawai, 2008),. as a tool for systematic study of classroom. a mere 0.012 percent of the British National. discourse, concentrating mainly on interactions. Corpus. However, because the classroom discourse. between the teacher and individual students.. environment is composed of multiple speakers. Farooq (2000) devised a useful diagrammatic. seeking a common goal, and because abundant. representation of the IRF model. Ohta (2007), in. non-verbal behavior occurs in class, classroom. an example of quantitative analysis, showed that. corpora might reveal discourse features not. the overwhelming majority of classroom follow-. readily available in corpora based on unstructured. up moves are spoken by the teacher, that. conversations such as telephone calls between. learners have limited opportunities to directly. friends and family. As teachers rarely analyze. follow-up, and that learners are often forced to. themselves (supervisor critique being more. experience the teachers’ moves as peripheral. common), classroom corpora displaying the. participants. Regardless of what happens. exchange structure are hoped to provide. frequently in the classroom, Kasper (2001) says. opportunities for more thorough self-inspection. teachers can (at least in theory) actively. by reviewing what discourse elements resulted. involve learners in interactions while adhering. in communicative student responses. Students’. to the IRF pattern (note that the teaching. acquisition of discourse skills are hoped to be. technique assumed by the IRF model is. studied by analyzing the students’ discourse. unchallenged).. skills displayed to both the teacher and the. Walsh (2002, 2006) proposed a Self-Evaluation. peers.. of Teacher Talk (SETT) procedure by observing IRF recurrent patterns. McCarthy and Walsh. 1.2 Classroom discourse structure. (2003) expanded the coverage of the IRF model. The initiation, response, and follow-up (IRF). to non-teaching situations.. model (Sinclair & Coulthard, 1975), summarized. Due to the long-term popularity and influence. in Tables 1 and 2, has influenced classroom. of the IRF model, it has been considered as one. discourse analyses for many years.. of the most probable approaches to classroom. McCarthy (1991) demonstrated the IRF. exchange structure analyses (O’Keeffe et al,. model’s usefulness in analyzing patterns of. 2007).. interaction. Sinclair and Coulthard (1992). The IRF model has received criticism over. developed a tool for systematically studying. the years. Willis (1992) points out that some. 2.

(4) Designing an XML schema for classroom discourse visual representation through XSLT. discourse categories do not mesh with reality.. other than classroom discourse (McCarthy &. We believe this is an inherent limitation. Walsh, 2003).. stemming from implicit assumptions the IRF. However the IRF model’s full potential is yet. model makes with regards to teaching. to be unleashed due to the following reasons:. techniques where teachers control and directly. (e) disproportionate research interest in. participate in discourse with students (as. initiation and response (i.e., follow-up neglected). opposed to discourse occurring among students,. (Coulthard, 1992), (f) lack of discourse analysis. as found in peer-based or autonomous learning).. methods to study the full spectrum of linguistic. Coulthard and Brazil (1992) acknowledge the. communication (Malouf, 2001), (g) variability. non-verbal element of discourse. A vexing issue. and unreliability of transcriptions (Atkins,. for them was the “directing move”, which often. 2001), (h) difficulty in transcribing non-verbal. depended on the non-verbal messages. More. behavior (Brazil, 1997), and (i) practical. general reservations were voiced by Malouf. obstacles to collecting and transcribing suitable. (2001), who complained that most discourse. amounts of data (O’Keeffe, McCarthy, & Carter,. analysis methods assume dialogue, whereas. 2007). If these problems were overcome. classroom discourse involves far more than two. (perhaps partly by aligning transcriptions with. participants. Although not a criticism of the IRF. nonverbal cues appearing in audio and video. model per se, the IRF model can be interpreted. recordings), teachers could utilize the IRF. as a series of dialogues because the model. model for self-reflection and improvement.. assumes teacher vs. (multi-)student interaction, with the modification that one of the two. 1.3 XML representation. interactants (i.e., the student) changes frequently.. Based upon the discussions in the preceding. The IRF model needs to expand into or be. sections, this study aims to utilize the extensible. subsumed within models of multi-interactant. mark-up language (XML) to visually represent. discourse. This and other issues have led Hall. the classroom discourse structure that embodies. and Walsh (2002) to admit that the IRF model’s. the IRF interactions. XML has advantages to. potential value for teacher-student interaction. achieve this task due to its nesting structure. research has not been fully realized.. being similar to the IRF hierarchical. In summary, the main advantages of the IRF. characteristic.. model are as follows: (a) the IRF model is a. XML can describe self-explanatory. systematic classroom discourse analysis tool. documents called instances. The instances are. particularly suited for teacher-student. composed of sets of elements that can be. interactions where the teacher holds authority. represented by start tags followed by end tags.. and responsibility (Malouf, 2001; McCarthy &. Each element has its element name with. W a l s h , 2 0 0 3 ) , (b) t h e I R F m o d e l c a n b e. attributes. We can declare whether attributes. graphically represented (Farooq, 2000), (c) the. are required or optional. Elements can have. IRF model provides a framework that enables. contents, but it is not necessary for them have. teachers to involve their learners in classroom. to have one. When such cases occur, they are called. interactions (Kasper, 2001), and (d) the IRF. empty elements. Figure 1 illustrates a simple. model organizes some kinds of conversations. graphic representation of an XML element.. 3.

(5) KATAGIRI Noriaki and KAWAI Goh. element. attribute. attribute value. content. element: parent. element: child 1. Figure 1．XML element structure. An oval shape represents an XML element horizontally followed by an attribute marked by a curly rectangle. The upper right rectangle (next to the attribute) shows the attribute value.. element: child 2. element: child 3. Figure 3．Tree diagram of parent and child elements.. 2. Materials and methods 2.1 Designing an XML schema. Figure 2 represents an MXL instance resulted. We expanded and expressed Sinclair and. from the XML element structure designed in. Coulthard’s discourse representation (1975). It. Figure 1. XML tagging system, where elements. is a hierarchical model of classroom discourse.. are parenthesized by a start tag “<tag>” and an. We could design an XML schema thanks to its. end tag “</tag>,” which enables elements to. hierarchical structure. We aligned the classroom. have superordinate elements (called parent. discourse elements, attributes and values. nodes), subordinate elements (child nodes), and. (Tables 1 and 2), designed the structure into an. horizontal elements (sibling nodes). This. XML parent-child node structure by (a). parent-child-sibling structure makes XML. creating a “classroom corpus” root element. instances look like a tree diagram (Figure 3).. above the “lesson” element (thus allowing a course to consist of multiple lessons), and (b). <element attribute name=”attribute”> content </element> Figure 2．XML instance representation. The attribute is embedded in the start tag whose “element” name is “element.” The attribute has a name “attribute.”. adding attributes to some elements to define variable values. The XML elements and their attributes form the following tree structure (Figure 4). The teachers hold authority and responsibility for multi-interactant exchanges. Elements are shown in shaded rectangles, and their attributes. Table 1 IRF Elements, Attributes, and Value Samples Hierarchy high. low. Element classroom corpus lesson transaction exchange move act. Attribute (value sample) none id (1, 2, …), date (2015-09-15), subject (Lesson 1), instructor (JT1, ...) id, transaction type (preliminary, medial, terminal) id, exchange type (boundary, teaching) id, language (TL［default value］, L1), speaker (JT1), move typea (frame, …) act typeb (starter, elicitation, directive, bid, reply, comment, cue, …). Note. The top element “classroom corpus” is the root element, and does not have any attributes. For the other five elements, attributes describe values of variables, such as who is speaking (“speaker”), the phase of discourse (“transaction type”), and whether teaching is occurring (“exchange type”). The “act” element is the lowest element, and has text nodes that include utterance transcriptions. abThe complete samples of move type and act type are detailed in Table 2.. 4.

(6) Designing an XML schema for classroom discourse visual representation through XSLT. Table 2 Complete Value Samples of Move Type and Act Type Move type framing focusing opening answering follow-up. Act type marker starter elicitation check directive informative prompt clue cue bid nomination acknowledge reply react comment accept evaluate silent stress meta-statement conclusion loop aside othersa. Note. a”others” can entail any unexpected attribute that is observed during on-going classroom observation.. 22 attributes of the “act” element. Sinclair and Coulthard (1975) added “others” to encompass any unexpected attribute in the act element. Sinclair and Coulthard (1975) added “others” to encompass any unexpected attribute in the element act to enhance its application. The core of the model is the “move” element. There are five types of “move” elements. They typically occur in chronological order: A “framing” move marks the beginning of an exchange, followed by a “focusing” move where teachers state what will happen in the subsequent actual learning behavior, followed by a sequence of IRF move cycles that are shown in the figure as “opening”, “answering”, and “follow-up” moves. These elements, lesson, transaction, exchange, move and act, will form a nesting structure in an XML instance. We designed the nesting structure through designing an XML schema. Table 3 depicts a design sample to illustrate a nesting structure. See Appendices A through D for the complete XML schema. 2.2 Visualization of the IRF model We will briefly examine how a classroom teacher-student interaction is transcribed. Generally, classroom discourse transcripts align teacher utterances in one column (usually left because mostly it is the teacher that initiates a discourse), and student utterances on the other (usually right). Table 4 shows such a transcription of a classroom discourse. We will use this teacher-student interaction transcript in this study.. Figure 4．The IRF (initiation, response, and followup) model, the root element excluded.. We designed an XSL Transformations (XSLT) style sheet to transform the annotated XML source (a transcript expressed in an XML. in rounded rectangles. The figure shows all. instance according the the XML schema. elements and attributes except for 19 out of the. explained in Table 3) to an XML representation. 5.

(7) KATAGIRI Noriaki and KAWAI Goh. Table 3 XML Schema Sample to Construct a Nesting Structure of Elements 1 <xs:element name="move" xmlns:xs="http://www.w3.org/2001/XMLSchema"> 2 <xs:complexType mixed="true" > 3 <xs:choice minOccurs="1" maxOccurs="unbounded" > 4 <xs:element ref="act" /> 5 </xs:choice> 6・・・・・・・・・・・・・ 7 <xs:attribute name="move_type" use="required" > 8 <xs:simpleType > 9 <xs:restriction base="xs:string" > 10 <xs:enumeration value="framing" /> 11 <xs:enumeration value="focusing" /> 12 <xs:enumeration value="opening" /> 13 <xs:enumeration value="answering" /> 14 <xs:enumeration value="follow-up" /> 15 </xs:restriction> 16 </xs:simpleType> 16 </xs:attribute> 17 </xs:complexType> 18 </xs:element> 19<xs:element name="act" > 20 <xs:complexType > Note. “move” and “act” in bold represent XML elements. The element “move” in line 1 has a child element “act,” which is referred to by the element name “act” in line 19. The “move” element has an attribute named “move_ type” (line 7), which is represented the either one of the values expressed in the form of string (line 9) such as “framing”(line 10) and “follow-up” (line 14). Dots in line 6 represent omitted elements due to their irrelevance to demonstrate the nesting structure. Table 4 Sample Transcript Line 1 2 3 4 5 6 7 8 9 10 11 12. Good morning, everyone.. Teacher. Students Good morning, Mr. X.. How are you today?. I’m fine, too, thank you. Well, today, we are going to read the first section of lesson one. Turn to page ten, and let’s practice pronouncing new words. OK? Good.. I’m fine, thank you. And you?. Yes, Mr. X.. suited for preliminary discourse analysis.. depending on what we discover in the data.. Our style sheet extracts discourse element. Table 5 explains an XSLT style sheet output. contents and presents them in tabular form. design. The top two rows show the column. (Table 5).. headers of our XSLT style sheet output. The. We may add, delete, or merge XML attributes. very top row shows elements of the discourse. 6.

(8) Designing an XML schema for classroom discourse visual representation through XSLT. Table 5 Tabular Column Headers of Transformed XML Annotations and Discourse Hierarchy Representation Sample Transaction id Type 1 preliminary. 2. medial. id 1. 1. Exchange Type. Move. Type. Act Transcript. elicitation. Good morning, everyone.. reply. Good morning, Mr. X.. elicitation. How are you today?. reply. I’m fine, thank you.. elicitation. And you?. reply. I’m fine, too, thank you.. marker. Well,. JT1. metastatement marker. JT1. directive. STs JT1. check react comment. we’re going to read the first section of lesson one. So, turn to page ten, and let’s practice pronouncing new words. OK? Yes, Mr. X. Good.. id. Type. Speaker. 1. opening (I). JT1. 2. answering (R). STs. 3. opening (I). JT1. 4. answering (R). STs. 5. follow-up (F). STs. 6. answering (R). JT1. 1. framing. JT1. 2. focusing. JT1. 3. framing. boundary. boundary. 4a. 5 6. opening answering follow-up. Note. (I)=initiation, (R)=response. They are not shown in the XSLT output. JT= Japanese teacher, STs=students. Speaker names usually have to only represent either a teacher or student/s and therefore are anonymous. aThis opening move has two acts for its child nodes. The other moves have one act for their child node.. hierarchy from the left (the highest) to the. that in a cascading way leads us to the. right (the lowest), and the second row shows. exchange id=1, type “boundary” followed by the. attributes to be extracted from the element. move whose id is “1” and type is “opening”. above each column. “Transcript,” which is. uttered by “JT1” until the act transcript with its. transcribed utterance forms the text element of. type. I-R sequences are observed in move ids. Act, is not an attribute of the “Act” element, but. 1-2, 3-4, and 5-6. One move may not constitute. a content of the “Act element which is simply. one sentence nor utterance since it is classified. expressed “<act act_type> Transcript </act>”.. as different moves based on roles they play in. For example, the discourse hierarchy is. discourse. In the sample below, move id 1 +. represented by the headers from left to right.. move id 2 make one utterance of JT1 “Well, we’re. The table starts with the transaction whose id. going to read the first section of lesson one.” This. and type are “1” and “preliminary” respectively. tabulation represents a classroom discourse. 7.

(9) KATAGIRI Noriaki and KAWAI Goh. table including an IRF sequence, and makes it. (1975).. easy for human eyes to examine classroom. Figure 5 shows an XSLT style sheet output. discourse hierarchy and speakers’ moves, acts,. in html. Although the transcript is only the. and speeches. See Appendix F for the complete. initial segment of a lesson, we have extracted an. XSLT style sheet.. expected output we planned in Section 2 (Tables 3 and 5) through our XSLT style sheet. The output shows the bottom 4 ranks in the. 3. Results. classroom discourse hierarchical structure;. We will look at the results obtained through. transaction/exchange/move/act. Each element. our XML instance, namely a classroom corpus,. displays attributes such as “id”, “type”, and. annotated based on an XML schema that shows. “speaker”. When we read the table from left to. a classroom discourse structure, and an XSLT. right, from the top to the bottom in a cascading. style sheet, which generates a hierarchical. way until the Act element, where utterances. discourse structure with classroom discourse. are transcribed, we can grasp the classroom. elements proposed by Sinclair and Coulthard. discourse hierarchy.. Figure 5．XSLT output of the annotated classroom discourse transcript in html.. 8.

(10) Designing an XML schema for classroom discourse visual representation through XSLT. 4. Discussion and conclusion. set of the XML schema and the XSLT style sheet.. 4.1 Visual classroom discourse representation. First, the elements and attributes in our. We proposed a method that utilized an XML. classroom discourse XML schema may not fully. schema to represent hierarchical elements of. schematize every probable classroom speech. classroom discourse elements in XML. behaviors. The model’s choice of attributes. (Appendices A through D). The XML element. reveals that the model implicitly assumes a. nesting structure is particularly suited for. particular style of teaching, namely a teacher-. annotating discourse structure and variable. controlled cycle of issuing directives, executing. attributes. Our XSLT style sheet visually. the directives, and evaluating the outcome. As. produces classroom speech transcript annotated. we continue annotating the classroom speech,. with classroom discourse hierarchical structure. we may encounter “other” attributes that. elements (Figure 5). Our XML schema design. Sinclair and Coulthard (1975) mentioned. We. and XSLT output have shown the potential of. will need to add other elements and attributes. the XML schema for accumulating classroom. in order to clarify the classroom discourse. transcripts, namely building classroom corpora.. structure.. Such corpora can be utilized multi-purposefully. Second, we may have included unnecessary. thanks to variable XSLT output representations.. elements and attributes in our XML schema.. When we complete annotating the entire. Some do not need all the attributes that we. transcript of one lesson, we will have a complete. proposed. Others may need more attributes. classroom discourse structure table as designed. than we designed. If researchers can customize. in Tables 1 and 2. Transcripts with discourse. the schema when planning to build a classroom. annotation will help language teachers reflect. corpus, they will be more likely to obtain data. upon their speech and improve their classroom. they aim at. Declaring attribute groups in. interaction skills. We hope to provide to. different namespaces enables researchers to. language teachers as well as teacher training. import specific attribute groups according to. program organizers persuasive, convincing, and. their research purposes. They may contribute. educational corpora expressed in our classroom. other attribute groups to the namespaces in. discourse XML schema. Insofar as the. order for the other researchers to employ and. annotations are clearly defined and consistently. create a new schema.. applied, the resulting analyses are less likely to spark impressionistic debate that too often. 4.3 Future research. characterizes teaching critique. Experienced. We are currently annotating a full transcription. teachers might collect and analyze discourse. of classroom spoken data using our XML. data for self-inspection. Prospective teachers. classroom discourse schema. When we are. might consciously organize lessons as a series of. finished with annotation, we will examine. transaction-exchange-move-act units.. whether (1) we will be able to combine several transcripts to verify accumulating the annotated. 4.2 Limitations. data, and (2) we will be able to retrieve specific. We recognize at least two limitations in our. elements, for example extract nothing but. 9.

(11) KATAGIRI Noriaki and KAWAI Goh. speech transcripts without annotation. Although we are currently focusing on spoken language, we will soon transcribe nonverbal behavior such as gaze directions, facial. Katagiri, N., & Kawai, G. (2008). Lexical analyses of native and non-native English language instructor speech based on a six-month co-taught classroom video corpus. Proceedings of Interspeech 2008, 19841987.. expressions, and gestures of the upper body. We. Kipp, M. (2001). Anvil - A Generic Annotation Tool for. expect that verbal and non-verbal annotations. Multimodal Dialogue. Proceedings of the 7th European. will be time-aligned with audio and video. Conference on Speech Communication and Technology (Eurospeech), 1367-1370.. recordings using methods such as Kipp (2001).. Malouf, R. (2001). Towards an analysis of multi-party. We hope our XML schema will intrigue. discourse. Retrieved from http://citeseerx.ist.psu.edu/. classroom discourse researchers for developing. viewdoc/download;jsessionid=DE787B73302670FFA39. better schema design and eventually lead to betterment of presvervice and in-service language teacher education.. 3A0D4D52BFF56?doi=10.1.1.8.4024&rep=rep1&type= pdf McCarthy, M. (1991). Discourse Analysis for Language Teachers. Cambridge. Cambridge University Press. McCarthy, M., & Walsh, S. (2003). Discourse. In D. Nunan. (Ed), Practical English Language Teaching. REFERENCES. (pp. 173-195). New York, NY: McGraw-Hill. Ohta, A. S. (2007). Second Language Acquisition Processes. Atkins, A. (2001). Sinclair and Coulthard’s ‘IRF’ model. in the Classroom. In A. O’Keeffe et al. (Eds.), From. in a one-to-one classroom analysis. Retrieved from. Corpus to Classroom (p. 226). Cambridge Cambridge. http:// www.cels.bham.ac.uk/resources/essays/ Atkins_4.pdf Brazil, D. C. (1994). Pronunciation for Advanced. Language Teaching Library O’Keeffe, A., McCarthy, M. & Carter, R. (2007). From corpus to classroom: language use and language. Learners of English－Teacher’s Book. Cambridge:. teaching. Cambridge: Cambridge Language Teaching. Cambridge University Press.. Library.. Brazil, D. C. (1995). Classroom and Spoken Discourse.. Sinclair, J. McH. & Coulthard, M. (1975). Towards an. Centre for English Language Studies. Birmingham:. analysis of discourse: the English used by teachers and. The University of Birmingham. Brazil, D. C. (1997). The communicative value of. pupils. Oxford: Oxford University Press. Sinclair, J. McH., & Coulthard, M. (1992). Towards an. intonation in English. Cambridge: Cambridge. analysis of discourse. In M. Coulthard (Ed.), Advances. University Press.. in spoken discourse analysis (pp. 1-34). London:. Coulthard, M., & Brazil, D. C. (1992). Exchange structure. In M. Coulthard (Ed.), Advances in spoken discourse analysis (pp. 50-78). London: Routledge. Farooq, M. (2000). Examining a Male Teacher’s Attention in a Mixed-sex EFL. Retrieved from http:// www. birmingham.ac.uk/documents/college-artslaw/ cels/essays/matefltesldissertations/fardiss.pdf. Routledge. Walsh, S. (2002). Construction or obstruction: teacher talk and learner involvement in the EFL classroom. Language Teaching Research 6(1), 3-23. Walsh, S. (2006). Talking the talk of the TESOL classroom. ELT Journal, 60(2), 133-141. Willis, D. (1992). Caught in the act: using the rank scale. Francis, G., & Hunston, S. (1992). Analysing everyday. to address problems of delicacy. In M. Coulthard. conversation. In M. Coulthard (Ed.), Advances in. (Ed.), Advances in spoken discourse analysis (pp. 111-. spoken discourse analysis (pp. 123-161). London:. 122). London: Routledge.. Routledge. Hall, J. K., & Walsh, M. (2002). TEACHER－STUDENT INTERACTION AND LANGUAGE LEARNING. Annual Review of Applied Linguistics, 22, 186-203. Kasper, G. (2001). Four Perspectives on L2 Pragmatic Development. Applied Linguistics, 22(4), 502-530.. 10.

(12) Designing an XML schema for classroom discourse visual representation through XSLT. Appendix A．Classroom corpus XML schema: schema/classroomCorpus/lesson/(transaction). <?xml version="1.0" encoding="Shift_JIS"?> <xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema"> <xs:element name="classroomCorpus" xmlns:xs="http://www.w3.org/2001/XMLSchema"> <xs:complexType xmlns:xs="http://www.w3.org/2001/XMLSchema"> <xs:choice minOccurs="1" maxOccurs="unbounded" > <xs:element ref="lesson" xmlns:xs="http://www.w3.org/2001/XMLSchema"/> </xs:choice> </xs:complexType> </xs:element> <xs:element name="lesson" > <xs:complexType > <xs:choice minOccurs="1" maxOccurs="unbounded" > <xs:element ref="transaction" /> </xs:choice> <xs:attribute name="id" type="xs:integer" use="optional" /> <xs:attribute name="date" type="xs:date" use="required" /> <xs:attribute name="subject" type="xs:string" use="required" /> <xs:attribute name="instructor" use="required" > <xs:simpleType > <xs:restriction base="xs:string" > <xs:enumeration value="ET1" /> <xs:enumeration value="ETn" /> <xs:enumeration value="JT1" /> <xs:enumeration value="JTn" /> </xs:restriction> </xs:simpleType> </xs:attribute> <xs:attribute name="n_of_students" type="xs:integer" use="required" /> <xs:attribute name="observer" type="xs:string" use="optional" /> </xs:complexType> </xs:element>. 11.

(13) KATAGIRI Noriaki and KAWAI Goh. Appendix B．Classroom corpus XML schema: (lesson)/transaction/exchange/(move). -------------------------- See Appendix A for ancestor nodes of the transaction node. -------------------------<xs:element name="transaction" > <xs:complexType > <xs:choice minOccurs="1" maxOccurs="unbounded"> <xs:element ref="exchange" /> </xs:choice> <xs:attribute name="id" type="xs:integer" use="required" /> <xs:attribute name="transaction_type" use="required" > <xs:simpleType > <xs:restriction base="xs:string" > <xs:enumeration value="preliminary" /> <xs:enumeration value="medial" /> <xs:enumeration value="terminal" /> </xs:restriction> </xs:simpleType> </xs:attribute> </xs:complexType> </xs:element> <xs:element name="exchange" > <xs:complexType > <xs:choice minOccurs="1" maxOccurs="unbounded" > <xs:element ref="move" /> </xs:choice> <xs:attribute name="id" type="xs:integer" use="required" /> <xs:attribute name="exchange_type" use="required" > <xs:simpleType > <xs:restriction base="xs:string" > <xs:enumeration value="boundary" /> <xs:enumeration value="teaching" /> </xs:restriction> </xs:simpleType> </xs:attribute> </xs:complexType> </xs:element>. 12.

(14) Designing an XML schema for classroom discourse visual representation through XSLT. Appendix C．Classroom corpus XML schema: /(exchange)/move/(act). ------------------------ See Appendices A and B for ancestor nodes of the move node. -----------------------<xs:element name="move" > <xs:complexType mixed="true" > <xs:choice minOccurs="1" maxOccurs="unbounded" > <xs:element ref="act" /> </xs:choice> <xs:attribute name="id" type="xs:integer" use="required" /> <xs:attribute default="TL" name="language" use="optional" > <xs:simpleType > <xs:restriction base="xs:string" > <xs:enumeration value="TL" /> <xs:enumeration value="L1" /> </xs:restriction> </xs:simpleType> </xs:attribute> <xs:attribute name="speaker" use="required" > <xs:simpleType > <xs:restriction base="xs:string" > <xs:enumeration value="ET1" /> <xs:enumeration value="JT1" /> <xs:enumeration value="ST" /> <xs:enumeration value="STs" /> <xs:enumeration value="others" /> </xs:restriction> </xs:simpleType> </xs:attribute> <xs:attribute name="move_type" use="required" > <xs:simpleType > <xs:restriction base="xs:string" > <xs:enumeration value="framing" /> <xs:enumeration value="focusing" /> <xs:enumeration value="opening" /> <xs:enumeration value="answering" /> <xs:enumeration value="follow-up" /> </xs:restriction> </xs:simpleType> </xs:attribute> </xs:complexType> </xs:element>. 13.

(15) KATAGIRI Noriaki and KAWAI Goh. Appendix D．Classroom corpus XML schema: /(move)/act. <xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema"> -------------------------See Appendices A, B, and C for ancestor nodes of the act node. ------------------------<xs:element name="act" > <xs:complexType > <xs:simpleContent > <xs:extension base="xs:string" > <xs:attribute name="act_type" use="required" > <xs:simpleType > <xs:restriction base="xs:string" > <xs:enumeration value="marker" /> <xs:enumeration value="starter" /> <xs:enumeration value="elicitation" /> <xs:enumeration value="check" /> <xs:enumeration value="directive" /> <xs:enumeration value="informative" /> <xs:enumeration value="prompt" /> <xs:enumeration value="clue" /> <xs:enumeration value="cue" /> <xs:enumeration value="bid" /> <xs:enumeration value="nomination" /> <xs:enumeration value="acknowledge" /> <xs:enumeration value="reply" /> <xs:enumeration value="react" /> <xs:enumeration value="comment" /> <xs:enumeration value="accept" /> <xs:enumeration value="evaluate" /> <xs:enumeration value="silent_stress" /> <xs:enumeration value="meta-statement" /> <xs:enumeration value="conclusion" /> <xs:enumeration value="loop" /> <xs:enumeration value="aside" /> <xs:enumeration value="others" /> </xs:restriction> </xs:simpleType> </xs:attribute> </xs:extension> </xs:simpleContent> </xs:complexType> </xs:element> </xs:schema>. 14.

(16) Designing an XML schema for classroom discourse visual representation through XSLT. Appendix E．XML instance expressed through XML schema in Appendices A through D.. <?xml version="1.0" encoding="UTF-8"?>  <classroomCorpus xsi:noNamespaceSchemaLocation="/Users/katagirinoriaki/Desktop/XML/XML_corpus_Schema/XML_I RF_Schema.xsd" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"> <lesson date="2015-09-17" subject="English Greetings" instructor="JT1" n_of_students="40"> <transaction id="1" transaction_type="preliminary"> <exchange id="1" exchange_type="boundary"> <move id="1" speaker="JT1" move_type="framing"><act act_type="starter">Good morning everyone.</act></move> <move id="2" speaker="STs" move_type="answering"><act act_type="reply">Good morning, Mr. X.</act></move> <move id="3" speaker="JT1" move_type="opening"><act act_type="elicitation">How are you today?</act></move> <move id="4" speaker="STs" move_type="answering"><act act_type="reply">I’m fine. Thank you.</act></move> <move id="5" speaker="STs" move_type="follow-up"><act act_type="elicitation">And you?</act></move> <move id="6" speaker="JT1" move_type="answering"><act act_type="reply">I’m fine, too. Thank you.</act></move> </exchange> </transaction> <transaction id="2" transaction_type="medial"> <exchange id="2" exchange_type="boundary"> <move id="1" speaker="JT1" move_type="framing"> <act act_type="marker">Well,</act> </move> <move id="2" speaker="JT1" move_type="focusing"> <act act_type="meta-statement">today, we are going to read the first section of lesson one.</act> </move> <move id="3" speaker="JT1" move_type="framing"> <act act_type="marker">So,</act> </move> <move id="4" speaker="JT1" move_type="opening"> <act act_type="directive">turn to page ten, and let’s practice pronouncing the new words.</act><act act_type="check">OK?</act> </move> <move id="5" speaker="STs" move_type="answering"> <act act_type="react">Yes, Mr. X. </act> </move> <move id="6" speaker="JT1" move_type="follow-up"><act act_type="comment">Good. </act> </move> </exchange> </transaction> </lesson> </classroomCorpus>. 15.

(17) KATAGIRI Noriaki and KAWAI Goh. Appendix F．XSLT style sheet. <?xml version="1.0" encoding="UTF-8"?>  <xsl:stylesheet version="1.0" xmlns:xsl="http://www.w3.org/1999/XSL/Transform"> <xsl:output method="xml" indent="yes" encoding="UTF-8" /> <xsl:template match="/" > <html> <head><title><xsl:value-of select="/classroomCorpus/lesson/@date"/></title></head> <body> <h1><xsl:value-of select="/classroomCorpus/lesson/@date"/></h1> <table border="1"> <tr><th colspan="2">Transaction</th><th colspan="2">Exchange</th><th colspan="3">Move</th> <thcolspan="2"> Act</th></tr> <tr><th>id</th><th>type</th><th>id</th><th>type</th><th>id</th><th>type</th><th>Speaker</th> <th>Acttype</th><th>Transcript</th></tr> <xsl:for-each select="/classroomCorpus/lesson/transaction"> <tr><td><xsl:value-of select="@id"/></td><td><xsl:value-of select="@transaction_type"/> </td><td></td><td></td> <td></td> <td></td> <td></td><td></td></tr> <xsl:for-each select="exchange"> <tr><td colspan="2" height="20"></td><td align="center" valign="top"><xsl:value-of select="@id"/></td> <td align="center" valign="top"><xsl:value-of select="@exchange_type"/></td><td colspan="5" height="20"> </td></tr> <xsl:for-each select="move"> <tr><td colspan="4" height="20"></td><td align="center" valign="top"><xsl:value-of select="@id"/></td> <td align="center" valign="top"><xsl:value-of select="@move_type"/></td> <td align="center" valign="top"><xsl:value-of select="@speaker"/></td><td colspan="2" height="20"></td> </tr> <xsl:for-each select="act"> <tr><td colspan="7" height="20"></td><td><xsl:value-of select="@act_type"/></td><td><xsl:value-of select="."/></td> </tr> </xsl:for-each> </xsl:for-each> </xsl:for-each> </xsl:for-each> </table> </body> </html> </xsl:template> </xsl:stylesheet>. (片桐徳昭旭川校講師) (河合剛北海道大学大学院准教授). 16.

(18)