User's manual - Revision history

Rjawor: /* Proportional statistics */

2017-04-06T21:11:16Z

‎Proportional statistics

Rjawor: /* Single words */

2017-04-06T21:10:30Z

‎Single words

Rjawor: /* Single words */

2017-04-06T21:09:20Z

‎Single words

Rjawor: /* Single words */

2017-03-30T22:19:30Z

‎Single words

Rjawor: /* Predefined statistics */

2017-03-30T22:05:30Z

‎Predefined statistics

Rjawor: /* Statistics */

2017-03-30T22:04:40Z

‎Statistics

Rjawor: /* Proportional statistics */

2017-03-30T22:00:20Z

‎Proportional statistics

Rjawor: /* Proportional statistics */

2017-03-30T21:56:44Z

‎Proportional statistics

Rjawor: /* Proportional statistics */

2017-03-30T21:56:22Z

‎Proportional statistics

Rjawor: /* Proportional statistics */

2017-03-30T21:49:04Z

‎Proportional statistics

@@ Line 162: / Line 162: @@
 In this case we take all words annotated as nouns and compute, how many nouns are furtherly tagged as singular, feminine. We obtain the following results:
-[[File:proportionalResults.png]]
+[[File:proportionalResults2.png]]
-You can narrow the search by language and epoque, just like any other set of results. In this case, however, you only get the numbers of word forms.
+You can narrow the search by language, epoque and individual documents, just like any other set of results. In this case, however, you only get the numbers of word forms.
 Please note that in proportional statistics you can also specify if counted words are in initial position in the sentence. You can, for example, compute the percentage of all feminine nouns in initial position in the sentence within the set of all nouns. In order to make it work, you put in the "noun" tag into the "main search criteria" and select "any" as the "position in sentence". As the "additional specific criteria" you enter the "feminine" tag and select "initial" as "position in the sentence". However, not all such combinations are correct. You can't, for example, compute the percentage of feminine nouns in non-initial position within all nouns in initial position. If you enter sentence position criteria which are conflicting in this way, the "position in sentence" in the "main search criteria" will be automatically set to "any".

@@ Line 146: / Line 146: @@
 [[File:SingleWordsFound.png]]
-At the top of the results page you can see a filter. With the help of the filter you may narrow your search results to only show words from documents in a specific language(s) and/or coming from a specified epoque(s). At all times you may return to the default filter settings by choosing the option "any", both for language and epoque.
+At the top of the results page you can see a filter. With the help of the filter you may narrow your search results to only show words from documents in a specific language(s) and/or coming from a specified epoque(s). Also, you can filter by individual documents. At all times you may return to the default filter settings by choosing the option "any", both for language and epoque.
 Below the filter you can see information about the total number of found words. Words are displayed in a table below in portions of 15 words per page. You can switch pages by using the buttons "first", "previous", "next" and "last".

@@ Line 144: / Line 144: @@
 Search results are presented as follows:
-[[File:singleWordsResults.png]]
+[[File:SingleWordsFound.png]]
 At the top of the results page you can see a filter. With the help of the filter you may narrow your search results to only show words from documents in a specific language(s) and/or coming from a specified epoque(s). At all times you may return to the default filter settings by choosing the option "any", both for language and epoque.

@@ Line 140: / Line 140: @@
 [[File:singleWords.png]]
-Words are searched by the criteria entered by the means of the large input field. In the above example, two criteria were selected: ''CVB (grammar)'' and ''VRB (POS)''. This search would return all the words from the database, which are annotated as CVB on the level grammar and as VRB on the level POS and possibly annotated also by other tags. In order to add a search criterion, start typing tag name in the input field, choose one from a pop-up list and press ENTER. The criterion will appear below the search field (as in the picture). To delete a criterion, just click on it. You may also specify, if the searched words should appear as the first word in the sentence or be located in the middle of the sentence. Simply use an appropriate button in the section "Position in sentence".
+Words are searched by the criteria entered by the means of the large input field. In the above example, two criteria were selected: ''CVB (grammar)'' and ''VRB (POS)''. This search would return all the words from the database, which are annotated as CVB on the level grammar and as VRB on the level POS and possibly annotated also by other tags. In order to add a search criterion, start typing tag name in the input field, choose one from a pop-up list and press ENTER. The criterion will appear below the search field (as in the picture). To delete a criterion, just click on it. You may also specify if the searched words should appear as the first word in the sentence or be located in the middle of the sentence. Simply use an appropriate button in the section "Position in sentence".
 Search results are presented as follows:

@@ Line 170: / Line 170: @@
 === Predefined statistics ===
-The Tagger system comes with a set of predefined statistics. If some statistics are often generated, it is convenient to have their search criteria stored. In order to add a new predefined statistic to the list, please contact the system administrator.
+The Tagger system comes with a set of predefined statistics. If some statistics are often generated, it is convenient to have their search criteria stored and generate them with a single click. In order to add a new predefined statistic to the list, please contact the system administrator.
 The predefined statistics list also includes two useful reports: words with "add info" and sentences with "add info". They are used to present words or sentences which were annotated on the "add info" level. This level is often used to enter comments about annotations to be read by other users.

← Older revision		Revision as of 22:04, 30 March 2017
Line 167:		Line 167:

	Please note that in proportional statistics you can also specify if counted words are in initial position in the sentence. You can, for example, compute the percentage of all feminine nouns in initial position in the sentence within the set of all nouns. In order to make it work, you put in the "noun" tag into the "main search criteria" and select "any" as the "position in sentence". As the "additional specific criteria" you enter the "feminine" tag and select "initial" as "position in the sentence". However, not all such combinations are correct. You can't, for example, compute the percentage of feminine nouns in non-initial position within all nouns in initial position. If you enter sentence position criteria which are conflicting in this way, the "position in sentence" in the "main search criteria" will be automatically set to "any".		Please note that in proportional statistics you can also specify if counted words are in initial position in the sentence. You can, for example, compute the percentage of all feminine nouns in initial position in the sentence within the set of all nouns. In order to make it work, you put in the "noun" tag into the "main search criteria" and select "any" as the "position in sentence". As the "additional specific criteria" you enter the "feminine" tag and select "initial" as "position in the sentence". However, not all such combinations are correct. You can't, for example, compute the percentage of feminine nouns in non-initial position within all nouns in initial position. If you enter sentence position criteria which are conflicting in this way, the "position in sentence" in the "main search criteria" will be automatically set to "any".
		+
		+	=== Predefined statistics ===
		+
		+	The Tagger system comes with a set of predefined statistics. If some statistics are often generated, it is convenient to have their search criteria stored. In order to add a new predefined statistic to the list, please contact the system administrator.
		+
		+	The predefined statistics list also includes two useful reports: words with "add info" and sentences with "add info". They are used to present words or sentences which were annotated on the "add info" level. This level is often used to enter comments about annotations to be read by other users.

	== System configuration ==		== System configuration ==

@@ Line 166: / Line 166: @@
 You can narrow the search by language and epoque, just like any other set of results. In this case, however, you only get the numbers of word forms.
-Please note that in proportional statistics you can also specify if counted words are in initial position in the sentence. You can, for example, compute the percentage of all feminine nouns in initial position in the sentence within the set of all nouns. In order to make it work, you put in the noun tag into the first field and select "any" as the position in the sentence. In the second field you enter the feminine tag and select "initial" as position in the sentence. However, not all such combinations are correct. You can't, for example, compute the percentage of feminine nouns in non-initial position within all nouns in initial position. If you enter sentence position criteria which are conflicting, the criterion for position in sentence in the general category will be set to "any".
+Please note that in proportional statistics you can also specify if counted words are in initial position in the sentence. You can, for example, compute the percentage of all feminine nouns in initial position in the sentence within the set of all nouns. In order to make it work, you put in the "noun" tag into the "main search criteria" and select "any" as the "position in sentence". As the "additional specific criteria" you enter the "feminine" tag and select "initial" as "position in the sentence". However, not all such combinations are correct. You can't, for example, compute the percentage of feminine nouns in non-initial position within all nouns in initial position. If you enter sentence position criteria which are conflicting in this way, the "position in sentence" in the "main search criteria" will be automatically set to "any".
 == System configuration ==

← Older revision		Revision as of 21:56, 30 March 2017
Line 165:		Line 165:

	You can narrow the search by language and epoque, just like any other set of results. In this case, however, you only get the numbers of word forms.		You can narrow the search by language and epoque, just like any other set of results. In this case, however, you only get the numbers of word forms.
		+
		+	Please note that in proportional statistics you can also specify, if counted words are in initial position in the sentence. You can, for example, compute the percentage of all feminine nouns in initial position in the sentence within the set of all nouns. In order to make it work, you put in the noun tag into the first field and select "any" as the position in the sentence. In the second field you enter the feminine tag and select "initial" as position in the sentence. However, not all such combinations are correct. You can't, for example, compute the percentage of feminine nouns in non-initial position within all nouns in initial position. If you enter sentence position criteria which are conflicting, the criterion for position in sentence in the general category will be set to "any".

	== System configuration ==		== System configuration ==

← Older revision		Revision as of 21:49, 30 March 2017
Line 155:		Line 155:

	=== Proportional statistics ===		=== Proportional statistics ===
		+
		+	With the help of proportional statistics, you can compute the percentage of specific word forms in a set of all word forms of the same type. Consider the following example:
		+
		+	[[File:proportional.png]]
		+
		+	In this case we take all words annotated as nouns and compute, how many nouns are furtherly tagged as singular, feminine. We obtain the following results:
		+
		+	[[File:proportionalResults.png]]
		+
		+	You can narrow the search by language and epoque, just like any other set of results. In this case, however, you only get the numbers of word forms.

	== System configuration ==		== System configuration ==