KanjiLab User Guide

Beta Release 0.5.2 - June 9, 2004

Collin McCulley (cmmcculley@charter.net)

This is a beta version, i.e. in need of user testing. If you find a bug or have a suggestion, please write to me at the above address and let me know. Thanks for using KanjiLab!

Acknowledgements

I'd like to thank the following people:

• George Knoll, for helping formulate the original concept, for helping to set requirements, for beta testing, and for the concept of time-graduated review.
• Graham Leonard, for providing an alternate set of data files and helpful suggestions and comments.
• Randy Foreman, for beta testing and helpful suggestions and corrections, as well as the Google data for the examples sorting feature.
• Frej Bjon and George M. Kleinman for useful feedback.
• Hirofumi Nagamura, for advice on kanji readings.
• Bart Mathias, for advice on kanji readings.
• Jim Breen, for creating the JMDict and KanjiDic data and providing it to the world, as well as for maintaining and providing the Tanaka corpus example sentences.

What is KanjiLab?

KanjiLab is a new kind of kanji drilling software whose key difference from other flashcard-style drills is requiring readings to be typed instead of selected from a list or passively revealed. The user must also know which readings are on-yomi and which are kun-yomi. This is a much more certain way of ensuring that the user is fully comfortable with the readings, and provides an active tactile involvement in the study process.

The readings in KanjiLab also act as a nexus for data, allowing the user to see examples keyed to each particular reading, and okurigana and position information for the reading. Before attempting a reading, a user can see example compounds containing that reading. In this way, readings can be learned and remembered in the context of vocabulary, rather than in isolation. The example set has been chosen to include a large selection of commonly used words.

KanjiLab also provides contextual study in the form of a fill-in drill, which presents words drawn from the example set for which the user must supply the missing kanji reading.

KanjiLab provides mastery score tracking. As the user becomes more familiar with a kanji, and this score increases, the kanji appears in drilling less frequently. Conversely, kanji which need more work appear more frequently. A unique score-declining methodology degrades these scores over time, so that review is automatic. Each time the user re-masters the kanji, the score declines more slowly, making review less frequent.

Ideally this software should be used in conjunction with a kanji study regime, such as provided by the Bojinsha series of books ("Basic Kanji Book" vol. 1 & 2, and "Intermediate Kanji Book" vol. 1 & 2), or "Kanji in Context". A configurable study list lets you introduce kanji in the order you are studying them in your primary learning environment.

Primary Features

• Shows you the exact number of readings of each type that are expected, and reveals each one as you get it correct.
• Uses a scoring system to keep track of your mastery of kanji, presenting the kanji you need to work on more frequently.
• When you select a revealed reading, full data showing okurigana and the word positioning is shown for that reading.
• When you click a revealed reading, you can cycle through a set of example compounds which use that particular reading.
• When you click on an unrevealed reading, you can see the example compounds without the reading and gloss, to give you a contextual hint.
• Optionally you can hide the meanings of the kanji until you click on them (on by default).
• Allows you to create a study list, listing the kanji in the order you want them introduced into the active set. Comes with a Jouyou grade ordered list as well as one for the Bojinsha kanji books.
• Fill-in drill provides practice with kanji in the context of commonly used words.

Planned Functionality

Two more drills will be added which are multiple-choice reversals of the readings and fill-in drills. The reverse readings drill will present a set of readings to the user, and have the user select from a set of kanji. The reverse fill-in drill will present a compound with the target kanji missing and a set of kanji which all possess the target reading.

Work on the kanji and examples data is always ongoing. Users should check back on http://www.epochrypha.com/japanese/ for periodic data (and program) updates.

Future versions will also have the following improvements:

• Support for multiple users.
• On-screen editing of kanji and examples data.
• On-screen editing of user kanji data (scores, active set, etc.)

Disclaimers

This software is provided AS IS. NO WARRANTIES are granted, either expressed or implied, and NO GUARANTEES are made of merchantability or fitness for a particular purpose. The author shall not be liable for any special, incidental, consequential or punitive damages arising out of the installation, performance or use of this software.

The author makes no claim of expertise in Japanese language pedagogy.

Copyright

This software is COPYRIGHT (c) 2003 by Collin McCulley.

This current version is provided free of charge and may be freely distributed by private individuals by any means so long as no money is charged, but must remain in its original form and must be accompanied by this document. It may not be included in freeware/shareware archives of any type or be distributed by commercial enterprises without the author's written permission.

The information derived from JMDict and KanjiDic sources is used in accordance with the public-use license from the Electronic Dictionary Research and Development Group, Monash University, Clayton, Victoria, Australia.

Menus

Drill Menu

Readings	Displays the readings drill.
Fill In	Displays the fill-in drill.
Exit	Quit KanjiLab.

User Menu

Options

Displays a dialog of user-configurable options.

Kanji Menu

Introduce	Forces introduction of kanji from the study list into the active set. By default, the number of kanji introduced at a time is 10. Also by default, the auto-introduce feature is turned on, which introduces new kanji once a certain percentage of the active set (default 75%) is mastered. These defaults can be changed by selection Options from the User menu.
Show Notes	Opens the notes window. This window may be left open while working in the drill window, and will change for each kanji. Notes are automatically saved as you move from kanji to kanji, and are written to disk when you exit the program. The notes window supports Japanese text, and will work with the IME, allowing about 1000 unicode characters. Be careful to switch the IME back to ascii input after typing, because it will interfere with keyboard processing in the drill window.
Decrease Mastery Level	Decreases the mastery level of the currently shown kanji by one, which will cause review to be more frequent.

Readings Drill

The basic goal of the readings drill is the memorization of 'on' and 'kun' readings associated with each kanji on your study list. This is done in conjunction with context provided by examples which use those readings. This context is important to remembering the readings, and is more effective than memorization in isolation.

First you should study the readings and examples together. In particular, look for words you already know. If none, pick a word or two you can use as a hook to remember the reading by (for instance, "this is the 'kan' in 'eigakan'"). Then when drilling, click on an unrevealed reading to see the example compounds. Use your memory of those words to help you remember the reading. For a greater challenge in later sessions, you can think of words you know which use the kanji without looking at the examples and use your memory of those words to remember as many readings as you can.

The probability that a kanji in the active set is presented to you in the drill is based on the mastery score of that kanji. You will see lower-scored kanji more frequently than higher-scored ones. By default, you will not see mastered kanji until the score decline mechanism brings the score back below the mastery threshold (default 90%).

The following are the parts of the readings drill window and how each works:

Kanji Display
Shows the target kanji in the 48x48 font. The Jouyou grade level of the kanji is displayed below it for grades 1-6, and a further divsion for secondary school kanji, grade 8. For non-Jouyou kanji, other numbers may be used to form groups.

Meanings
Shows the meaning of the kanji in English. By default, this is hidden until you click on it or press F3, so you can use this as a flash card to learn the meanings, or use the meanings as a hint for remembering the readings. On the options dialog, you can choose to have the meanings always stay revealed.

Kana Type-In Field
This field is where you will type in your attempts. It works similarly to an IME. Simply type the roomaji and it will be converted to kana. If you are currently set to attempt 'on' readings, you will see katakana. If you are attempting 'kun' readings, you will see hiragana. Press ENTER to complete your attempt. If you leave a syllabic "n" unconverted on the end of what you have typed, it will convert to kana automatically, so there is no need to type "n'" unless you need to prevent a following vowel or "y" from being misinterpreted. You can type an 'x' before 'tsu' or 'tu' to get a trailing small tsu on those readings which require it, but it is just as easy to type a double consonant like 'kk' or 'tt'. This will leave an unconverted letter in the buffer, but that will not affect the final conversion.
     The number of attempts remaining is shown to the right of the type-in area. You have three attempts per reading, but you may use more on some readings and less on others. You must simply get all the readings in the total number of attempts allowed. Note that if you type a reading which is a repeat of a previously missed attempt, it will not count against you as long as that attempt remains in the Incorrect Tries buffer (see below). If you run out of attempts, the remaining readings will be revealed, and the card will be scored. There is an option to allow unlimited attempts.

Score Information
This area shows you your current mastery score and mastery level (the scoring system is explained in detail below), and the number of kanji you currently have in your active set, as well as the percentage of these which are currently above the mastery level. After you complete a card, a box will appear to show you how the card was scored.

Readings
Each card is presented with a set of possible 'on' readings in the left column, and a set of 'kun' readings in the right column. The exact number you must guess is shown as unrevealed boxes in deep blue. At any one time, you are guessing either 'on' or 'kun' readings, which is indicated by a magenta box which surrounds the column you are currently working on. At any time you may change which column you are guessing in by pressing the TAB key. You may select a reading with the mouse or keyboard, in which case the current column will become the one from which the selection was made. The left and right arrow keys will scroll the reading selection, and for convenience, the comma and period keys behave the same way.
     When you select an unrevealed reading with the mouse, example compounds specific to that reading will appear in the Examples area (see below) without accompanying readings or glosses. When you make a correct guess, the corresponding reading's box becomes green, shows the reading and becomes the active selection. Selecting on a revealed reading will show okurigana and position information below the readings area (see below), and also examples for that reading with readings and glosses shown. If you give up and reveal answers on a card, or run out of attempts, the ones you did not get correct will appear in grey.

Incorrect Attempts
If you make an incorrect attempt, it will be shown in the area at the right of the window, so you may remember which attempts you have already made. If the number of incorrect attempts reaches the bottom of the screen (which is unlikely unless a kanji has a very large number of readings), older incorrect guesses will be replaced by newer ones.
     If you inadvertently try a previous incorrect attempt again, you are not penalized for it (i.e. it doesn't take away one of your remaining attempts) as long as the previous attempt is still in the buffer and displayed in this area.

Control Buttons
The following four buttons appear in the readings drill window:

• Next: Before starting to guess a card, you may choose to skip the kanji and go on to the next one. Once you have started to guess, you must complete the card or abandon it. The space bar is synonymous with this button. Note that if one kanji has a score far below the others in the active set (or is the only unmastered kanji in the active set), you may see it again when you press "Next".
• Reveal: If you wish to give up on a card before your allowable number of guesses is exhausted, this button will reveal the remaining readings. Your score is marked off for leaving readings unrevealed. Pressing this button has the same effect as running out of guesses. If the option for it is enabled, revealing a card or running out of guesses will lower the mastery level by 1 (if it was greater than 0).
• Mastered: This button is provided to allow you to mark as "mastered" kanji you are already familiar with, so you do not need to spend time drilling them, and instead only see them for review as the score decline mechanism brings them back into the drill set. Each time you press this button, the mastery level is increased by one. You may repeatedly press it to slow the rate at which the score will decline.
• Study: Before beginning to guess a card, you may choose instead to study it, which will reveal all readings and examples. This button becomes disabled once you start guessing on a card, at which point you have to click "Reveal" and take your lumps. If you study a card, you have to wait until the next time it comes up to make a scored attempt. It is recommended that if you know any reading on a card that you not use the study button, and instead just get the readings you can. Taking a low score on a card will help it come back for review more often, so you can master the remaining readings.

Examples
This area displays example compounds corresponding to a selected reading, with compound reading and gloss hidden if the reading is unrevealed. To scroll through the available examples, click on the example box, or use the up and down arrow keys. The left and right bracket keys are also aliases for this, so you don't have to move your hands from the main keyboard.
     When there are examples of compounds which use the kanji but not with one of the official readings, a "Special Readings" box will appear below the Examples area. Select this box to see those examples. The 0 key on the keypad will also select the special readings box.
     Examples are sorted based on a frequency score so that more common words appear first. These scores have been initially generated based on Google hit data. Larger scores indicate a more common word. These scores may not be perfect indicators, and can be edited to suit the user.
     Starting in version 0.5.0, the example sentences used with WWWJDIC (the "Tanaka corpus") have been tied into the example compound feature. To use this feature you must separately download the "sentences.dat" file and place it in the KanjiLab installation folder. When sentences are available for a compound, a "Sentences" button will appear below the examples area. Clicking this button will bring up a window with the associated sentences. Note that the Tanaka corpus is a work in progress, and has many errors, including errors in the Japanese, errors in the English (sometimes complete mistranslations) and errors in the indexing. Nevertheless, it is a valuable resource for students.
     If you find errors, they may be reported to Jim Breen, who maintains these examples. Use WWWEXAMPLES as the subject line, as this lets the messages be collected by script. Include both the Japanese and English, with your comments or corrections below each sentence or group of sentences to which your correction applies. You should update your copy of the sentences.dat file frequently or check the same sentences on WWWJDIC if you will be sending corrections so that you avoid repeating corrections that have already been made. (Note: There is an problem with cutting and pasting from the sentences window to a JIS formatted mail message in Outlook. A workaround is to first paste it into Notepad, then copy from Notepad to Outlook.)
     You can find more information about the sentence examples on the WWWJDIC site, including where to download a raw copy of the file. The only difference between the raw file and the one used for KanjiLab is the encoding and the name. If you take the raw file, use JWPce or another capable program to save the file as Unicode (UCS2), and rename it "sentences.dat", it will work with KanjiLab.

Okurigana/Position Info
This box displays the raw reading data for the selected reading. This includes such information as official okurigana in parentheses, and markings which indicate what position the reading may take in compounds.

Fill-In Drill

The goal of the fill-in drill is to provide practice with the readings you know in the context of commonly used words. Obviously, knowing readings is not enough. To read Japanese, you must also know what reading is used in a given compound. The selection mechanism for kanji in this drill is the same as for the readings drill. By default, mastered kanji are selectable in this drill.

Each card of the drill consists of a single word or phrase using the target kanji drawn from the examples set. The target kanji is highlighted in blue. The reading is presented below the word with the part represented by the target kanji left blank. You have three attempts to type in the missing reading. You must type in the reading as used in the word, which may include variations from the listed readings because of sokuon (e.g. shutsu->shupp in shuppin) or rendaku (e.g. hana->bana in hibana).

The kanji display, meanings, kana type-in field, score information area, and incorrect attempts list all work the same as in the readings drill, as do the next and reveal buttons (except that revealing a card or running out of guesses will NOT affect the mastery score). You may click the gloss area or press F4 to reveal the gloss at any time before completing the card. It is automatically revealed when the card is completed. The score for the fill-in drill is directly added to the mastery score for the target kanji, and is calculated as follows.

If correct:

                 3%
S =  --------------------------
     num_readings * num_guesses

and if incorrect:

         -3%
S =  ------------
     num_readings

Where num_readings is the total number of readings (on-yomi and kun-yomi) for the kanji. This scoring method reflects the fact that for a kanji with more readings, a single fill-in card explores less of the "readings space", and so moves the mastery score by a smaller amount. Scores in this drill can be fractional, so their effect on the mastery score may not be immediately apparant if less than 1%.

Reading Drill Scoring and Mastery Score Tracking

Each card of the reading drill is scored based on the percentage of readings you get correct minus a penalty for the number of attempts you take beyond one attempt per reading. A perfect score consists of getting all the readings correct without any misses. The overguess penalty, which in the score box appears subtracted from the percentage correct, is calculated as follows:

                   12.5
overguess =  ------------------  * (guesses - number_of_readings)
             number_of_readings

And now the math:

If a kanji card is completed with a percentage score of S, the new mastery_score is:

                (score_inertia - 1)                        1
mastery_score = ------------------- * mastery_score + ------------- * S
                   score_inertia                      score_inertia

                                          D         
mastery_score =  mastery_score  -  ----------------  * score_baseline
                                   mastery_lifespan

Kanji Selection

Kanji are selected for the drills based on a weighted roulette wheel approach. The size of a given kanji's slice of the roulette wheel (its selection weight) is based on the mastery score, with the effect that lower-scored kanji are selected more often. Selection weight is also a function of the selection distribution number which may be set by the user in the options dialog.

                          100 ^ selection_distribution
selection_weight =   -------------------------------------
                     mastery_score ^ selection_distribution

Where ^ indicates exponentiation. If the mastery score is fractional, the numerator only is used.

Lower values of the mastery score will result in a greater chance of selection. The magnitude of the difference caused by mastery score is scaled by the selection distribution parameter, such that higher values of the selection distribution will cause more bias towards less well known kanji, and lower values will flatten the distribution, equalizing the chance for selection.

The minimum selection distribution is 1.0, which will leave selection weight inversely proportional to the mastery score. In that case, a kanji with a mastery score of 50% will be 1.5 times more likely to be selected than a kanji with 75% mastery. For the default 4.5 selection distribution, a kanji with score 50% is about 6.2 times more likely than one at 75%.

Note that if the proportion of kanji in the drill set with higher scores is significantly greater than the number of lower scored kanji, higher scored kanji will still seem to appear pretty frequently. Individually, however, the probability is smaller.

The User Profile

The user profile "user_profile.klu" file is where KanjiLab stores user options and the user's history. This file is created the first time the program is run. It is a text file, and may be opened with Notepad or any other text editor. Be sure to save it as a text-only file. Now that KanjiLab has an options dialog, you should avoid hand-editing it. Make a backup copy if you do.

The first two lines at the top of the file should never be edited. The first line contains version information that will allow future versions of KanjiLab to read and update the format of this file. The second line is a machine-readable time for tracking score decline.

Following that are a number of user-settable parameters. Each has documentation in the file showing what it is (ones that don't shouldn't be touched).

After the parameter lines are the user kanji records. This is a tab-delimited record with the hexadecimal JIS code, a flag indicating if the kanji is in the active set (0 if not, 1 if so), the mastery score, and the mastery level. Note that kanji records are synchronized to the kanji data at program start up, so kanji missing from the kanji data will be deleted, and kanji in the data but not the user file will be added as baseline records.

Study List File Format

The study list is simply an ordered list of kanji in a Unicode UCS-2 file. It may have linebreaks, but no whitespace. Until a selection mechanism is built into the program, the program looks for a file in the same folder as the application with the name "study_list.kls". Simply copy your file into the folder and give it this name (you will need to delete or rename the default study list first).

The program is robust against the inclusion of kanji for which there is no kanji data on the study list. Such kanji will simply be skipped. KanjiLab will also skip any duplicate kanji which appear on the study list.

Kanji Data File Format

The kanji_data.dat file is a unicode ucs-2 file, which you can read with Notepad (with a Japanese font set), JWPce, etc. There is one line per kanji, with five fields per line. These fields are separated by tabs.

If you are adding non-Jouyou kanji, you can make the grade whatever you like, as long as it is an integer. The native data set uses grade 8 to indicate the highschool kanji. The jinmeiyou kanji, when they are added, will be grade 9.

In the readings sections, readings are separated by ideographic slashes. Within each slash division may be multiple forms to account for different okurigana or position markers, and these must be separated by ideographic spaces. The reading may be preceeded or followed by the position marker, and may be followed by okurigana inside ideographic parentheses. The actual reading must be identical for all cases inside a slash division.

Examples Data File Format

The examples.dat file is a unicode ucs-2 text file, which you can open with Notepad. Set the font to a Japanese font like Mincho.

Each entry has four tab-separated parts, and exists on a single line (Notepad may show a line break if you have word wrap turned on, but there is no actual carriage return).

The brackets around the reading must be the heavy style ideographic brackets.

The reading must be separated by IDEOGRAPHIC slashes, i.e. double-byte Japanese slashes. The number of divisions must exactly equal the number of characters (including kana and small kana) in the written form.

In order to show up in the program, the reading within a division must exactly match a reading that is part of the kanji data. However, this may not be the case because of sokuon (syllable converted to small tsu) or rendaku (change to a voiced reading). In these cases you must force the reading to associate to the one it originally came from. You do this by using IDEOGRAPHIC parentheses immediately following the reading and within the division, in which you put the reading to force association with.

If a reading in a division is irregular for the associated kanji, but readings in the compound may still be associated kanji by kanji, an ideographic ampersand should be placed before the irregular reading. For that kanji, the compound will show up as a "special reading" case. If you can't clearly associate any part of the word's reading with any of the individual kanji, i.e. it's totally irregular, you can put an ideographic asterisk before the reading, and skip making any divisions (i.e. DO NOT use any slashes when using the asterisk marker). The example will then show up in the "special readings" for all kanji in the compound.

In cases where a kanji posesses the same reading as both an on-yomi and kun-yomi, specific example readings can be forced to sort as one or the other. For instance, most two-kanji compounds would be considered to be using the on-yomi, whereas verb or adjective forms with okurigana would most likely be kun-yomi cases. An ideographic left-pointing angle bracket < is used to mark an on-yomi, and a right-pointing angle bracket > is used to mark a kun-yomi. This symbol immediately preceeds the reading, and may not be combined with & or any other symbol.

For the frequency score, higher numbers indicate a more common word, and will cause the word to sort closer to the top. The initial data was generated using Google hit scores, but any appropriate ranking may be substituted.

One last thing is very important. There can be no blank lines in the file, and that includes at the end. If you scroll to the end of the file you will see that you can't move the cursor past the last line and onto its own line. Be careful not to introduce any extra carriage return there, and be sure to delete any blank lines after you edit the file.

Alternate Data Set

You can get an alternate set of kanji data and examples constructed by Graham Leonard at:

http://www.kirisame.org/Kanji/

These were constructed to match the Kanji in Context workbooks.

Note that the format of the examples file has changed from version 0.4.0 to version 0.5.0. These changes may or may not have been included in this data set.

Changes New To Version 0.3.0 Beta

1. Fill-in drill completed. Three additional options related to this drill were added:
   • An option to not count the scores from the fill-in drill in mastery score totals.
   • An option to keep mastered kanji in the fill-in drill which is separate from that option for the readings drill.
   • An option to have the fill-in drill present only on-yomi or only kun-yomi cases (with the default of presenting all cases).
2. An option has been added, enabled by default, which causes the mastery score to jump to 100% once the mastery threshold has been crossed. This prevents a kanji from too quickly re-entering the drill pool because of score decline.
3. Space bar is now always "next", whether the button has focus or not. It's synonymous with F9.
4. The examples layout has been tightened up a little to allow more of a definition to show when it is long.
5. The percentage of the active set that is currently mastered is now shown in the drill window.
6. F3 will now reveal kanji meanings in the readings drill.
7. The left and right arrow keys will scroll through reading selections.
8. The zero key on the keypad will select the special readings.
9. The examples can now be scrolled with the up and down arrow keys.
10. A bug that caused problem with score decline if a date ahead of the current date was saved in the user profile has been fixed.
11. Additional corrections to the kanji readings.

Changes New To Version 0.3.1 Beta

1. Revised the default score decline schedule to half as fast.
2. Fixed a bug which caused repeated auto-introduce attempts when the study list was complete and the percent of cards mastered exceeded the auto-introduce threshold.
3. Fixed a bug which allowed negative mastery levels to occur if the user-profile was inconsistently hand-edited.
4. Added an options dialog selectable from the User menu, including the abililty to edit the score decline schedule.
5. A few minor example file corrections.
6. Fixed a bug which would prevent introduction of new kanji if the user switched to a study list smaller than the current active set.
7. Aliased the comma and period keys to scroll readings like the left and right arrow keys, and the left and right bracket keys to scroll examples like the down and up arrow keys. This way you don't have to remove your hands from the main keyboard for these activities (or strech and hurt your CTS-affected wrist, like I was doing). The arrow keys still work.

Changes New To Version 0.4.0 Beta

1. The examples data now allows for including examples where one or more individual kanji use an irregular reading, but where other kanji in the same compound use regular readings. The irregular readings are prefixed by an ideographic ampersand in the examples.dat file. On the kanji matching that position, the example shows up under "special readings", but for the other kanji it matches the regular reading. Many new examples have been added which take advantage of this capability.
2. Added an option to have the mastery level drop by one if the user presses reveal, or runs out of guesses on a readings drill card. Enabled by default.
3. Added an option for the fill-in drill to drill on mastered kanji only. This allows that drill to be used as a capstone exercise for the kanji the user is relatively familiar with. Off by default.
4. If a kanji has no on-yomi, then the readings drill will now start with focus on the kun-yomi column.
5. If a readings drill card is completed correctly, then the "mastered" button becomes re-enabled, so you can make the decision to use it after trying the kanji once.
6. Added an option to reverse the function of the space bar and enter key, so that space bar becomes "convert", and the enter key becomes "next". Off by default.
7. Fixed a bug which prevented kanji from being introduced according to the user's chosen score baseline. Before the pre-generated score from the user profile would be assigned.
8. Added the ability to keep user notes for each kanji. The notes window supports Japanese text and will work with the IME, allowing about 1000 unicode characters to be stored for each kanji. The notes window may be left open while working in the drill window.
9. Control has been asserted over the font size, so that running KanjiLab on machines where the default font size set to larger than normal should no longer cause problems for window layout (except that text on buttons may still be slightly clipped).
10. The "special readings" button now only appears when there are actually special readings examples available for a kanji.

Changes New To Version 0.5.0 Beta

1. Examples are now optionally tied to the Tanaka corpus sentences by including the "sentences.dat" file in the installation. If an example has sentences, a "sentences" button appears below the examples box. Clicking on it will bring up a window with the associated example sentences. Note that adding the sentences will cause an increase in startup loading time by about 20-30 seconds as the file index is built.
2. The example data has been enhanced with Google scores, and examples are now sorted from highest to lowest score, so that (theoretically) more common words appear before less common ones.
3. The user guide has been cast in HTML, and can now be pulled up in the user's default browser from inside the program under the Help menu.
4. An option has been added to reveal the kanji meanings when a drill card is completed. Enabled by default.
5. A "Decrease Mastery Level" command has been added under the Kanji menu to allow the user to bring down a mastery level on a kanji so that it will be reviewed more frequently.

Changes New To Version 0.5.2 Beta

1. There is now a way to mark example readings so as to force their interpretation as on-yomi or kun-yomi. This is used for those cases where there are matching on-yomi and kun-yomi for a given kanji.
2. Typing "nn" will now convert completely to kana 'n' rather than leaving "n" in the buffer.
3. All examples from the examples_pool file, including cases of implicit okurigana have been migrated to the examples set.

Default Settings

The following are the default settings as generated in the user profile when KanjiLab is run for the first time.

If you want to update your profile, do so only _after_ running KanjiLab 0.5.0 with your previous profile, as your profile must be converted to contain the new options (i.e. don't try to cut and paste the new ones into the file).

50	<--score baseline
90	<--mastery threshold
4	<--score inertia
4.50000	<--selection distribution control
1	<--use score decline
10	<--number of kanji to introduce at a time
1	<--use auto-introduce
75	<--auto-introduce threshold
0	<--Reverse space and enter key function
0	<--allow unlimited guesses
0	<--keep mastered kanji in readings drill rotation
0	<--meanings stay revealed
1	<--reveal kanji meanings on complete
1	<--mastery score jumps to 100% on crossing mastery threshold
1	<--reveal drops mastery level
1	<--keep mastered kanji in fill-in drill rotation
0	<--drill only mastered cards in fill-in drill
0	<--fill-in drill does not affect mastery score
0	<--restrict reading type in fill-in drill