SCoRE: Singapore Corpus of Research in Education

Here in this demo query package, we have selected a list of transcripts from different subject, level, and stream. The breakdown of the transcripts of 31 English-mediated lessons for the demo is as follows:

Subject	Level	Stream	Units	Lessons	Schools	Duration(hour)	Tokens
English	Pri 5	EM1	0	0	0	0	0
		EM2	1	5	1	3.9	18308
		EM3	0	0	0	0	0
	Sec 3	EXP	0	0	0	0	0
		NA	1	3	1	3.4	19567
		NT	0	0	0	0	0
		SPE	0	0	0	0	0
Maths	Pri 5	EM1	1	2	1	2.1	10868
		EM2	1	1	1	1.0	5457
		EM3	0	0	0	0	0
	Sec 3	EXP	0	0	0	0	0
		NA	1	2	1	1.2	12087
		NT	1	1	1	1.0	7779
		SPE	0	0	0	0	0
Science	Pri 5	EM1	1	4	1	3.2	16541
		EM2	0	0	0	0	0
		EM3	0	0	0	0	0
	Sec 3	EXP	0	0	0	0	0
		NA	1	6	1	2.3	21850
		NT	0	0	0	0	0
		SPE	0	0	0	0	0
Social	Pri 5	EM1	1	3	1	1.4	10633
		EM2	0	0	0	0	0
		EM3	0	0	0	0	0
	Sec 3	EXP	1	4	1	1.8	11731
		NA	0	0	0	0	0
		NT	0	0	0	0	0
		SPE	0	0	0	0	0
Total			10	31	10	21.3	134821

As showed in the table above, the corpus size of the demo package here is about 21 hours recording and 134821 tokens (words) in total. On average, the recording duration of each of the 31 transcripts/lessons is about 42 minutes; and the word count of each of the 31 transcripts/lessons is about 4400 tokens (words).

If you want to play the audio clips and video clips of your search results, you must check the option "Show results in turns" in the lower left corner of the query window when submitting your queries. Please be reminded that there are a few misalignments of audio clips due to the resampling of audio files. We are currently fixing this problem. For testing purpose, you can only play some of the audio/video clips of the search results in the result list. The audio icon

and video icon

displayed in your query result indicate the availability of the media.

For a better performance with audio and video played in your browser, we recommend the latest version of Internet Explorer. If you do want to use Mozilla Firefox, some media files may not be played automatically, as Firefox has limited capacity to play Windows Media files, and you may require changes to your system configuration. We recommend your IT administrator assist in correcting the problem (the proper Windows Media and Active X plugins need to be installed). Go to: http://forums.mozillazine.org/viewtopic.php?t=206213 and follow the steps exactly. After doing so, you should be able to view the Windows Media files via Firefox Browser.

When you browse the search result from the query, you may find that we have indexed all the events in the classrooms:

Table 2. Event Index in the Transcripts

Event	Symbol	Explanation
ENT001	%%	Background conversation that is inaudible
ENT002	##	Background noise
ENT003	CHORUS	Choral voices
ENT004	\$	Laughter
ENT005	\$\$	Extended Laughter
ENT006	[$]	Laughter Quality
ENT007	[V]	Verbatim Reading
ENT008	(O)	May or may not be talk
ENT009	( )	Ungotten talk
ENT010	( )	Extended ungotten talk
ENT010	( )	Extended ungotten talk
ENT010	( )	Extended ungotten talk ??
ENT011	<MS:	Generic Male Student Background voice initiated
ENT011	<MS:	Generic Male Student Background voice initiated
ENT012	<FS:	Generic Female Student Background voice initiated
ENT012	<FS:	Generic Female Student Background voice initiated
ENT013	<US:	Generic Unknown Student Background voice initiated
ENT013	<US:	Generic Male Student Background voice initiated
ENT014	<TV:	Generic Teacher Background voice initiated
ENT014	<TV:	Generic Teacher Background voice initiated
ENT015	>	End of background voice
ENT016	[	Overlap Onset
ENT017	]	Overlap Termination
ENT018	==	Latching - indicate end of word and start of new word
ENT019	=	No break in utterance
ENT020	(.)	Gap between utterances
ENT021	(..)	Extended gap between utterances
ENT022	:	Prolongation of immediate prior sound
ENT023	::	Longer prolongation of immediate prior sound
ENT023	:::	Longer prolongation of immediate prior sound ??
ENT024	-	Cut off
ENT025	?	Rising intonation
ENT026	.	Falling intonation
ENT027	\\	Falling intonation contour
ENT028	/	Rising intonation contour
ENT029		Continuing intonation
ENT030	^	Stress
ENT030	^^	Stress
ENT031	[E]	Code-Switching - English spoken
ENT032	[C]	Code-Switching - Chinese spoken
ENT033	[M]	Code-Switching - Malay spoken
ENT034	[H]	Code-Switching - Hindi spoken
ENT035	[T]	Code-Switching - Tamil spoken
ENT036	[U]	Code-Switching - Unknown language spoken
ENT999		All others

We have created a scheme to index the speakers in the transcripts. In this way, the corpus process and database query can be easier, and it can also help us to generate the speaker's profile across levels, streams, subjects, etc. in corpus query. In addition, this index can also enhance the anonymity of speakers when the corpus data is released to the public. Here is the list of speakers you can find in the transcripts:

Speaker Type	Speakers	Index	Remarks
Researcher	Researcher	RA	Each transcript has one researcher, and some may have more than one.
	RA, Coder, Interviewer...
Teacher	Teacher	TR	Each transcript has one teacher, and some may have more than one teacher.
	Teacher 1, 2...
Student	Student	ST	Each transcript has more than one student participated.
	Student 1, 2...
Class	Class	CS
	Group, Some students...
Audio System	VCD, DVD, CD, MP3 player...	AO	Audio played in the class.
Video System	VCD/DVD player, TV...	VO	Video played in the class.
PA System	Public announcements	PA	Public announcements while class in progress.
Others	All the others	OT	All the others in the class.

The linguistic features that you can find in the demo data are listed in the table below.