JPS6079421A - Selection system for homonym display - Google Patents

Abstract

PURPOSE:To perform efficiently selecting operation by shifting up and left a homonym group display when the display in a sub-window extends beyond the lower and right limits of the scope. CONSTITUTION:The keyboard 11 of a Japanese word processor includes functions keys such as a conversion key in addition to KANA (Japanese syllabary key) and also has a central processing unit 12, etc. KANA Information is inputted and stored in a character code buffer 21, and its character pattern is read out of a character generator 19, and written in a frame memory 17 and also displayed on the display scope 16. The 1st candidate of KANA-KANJI (Chinese character) conversion processing and the total number of homonyms from a module 13 are displayed in a main window on the scope 16. Further, when a next candidate key is operated, said operation is repeated to display a candidate group of homonyms in the sub-window. In this case, the display extends beyond the lower or right side limit of the scope, the display is shifted up or left.

Description

[Detailed description of the invention]

<Technical Field of the Invention> The present invention relates to a homophone display selection method in a text processing device such as the Japanese word processor Tsusa which employs a kana-kanji conversion method. <Technical background of the invention and its problems> As a method for displaying and selecting homophones in a conventional kana-kanji conversion type text processing device, only the first candidate homophone of the conversion result is displayed on the display screen. The first method (for example, Japanese Patent Application Laid-open No. 53-56925) updates the display to the next candidate homophone by operating a specific key, or displays all the homophones of ゛ on the display screen. , the second method of selecting a desired homophone from among them (for example, Japanese Patent Application Laid-open No. 5
3-56926) is known. However, in devices that adopt the conventional Juki 1st method, only one homophone is displayed as a result of Kana-Kanji conversion, so it is difficult to know how many key operations it takes to get the desired word. This poses a problem in that it places a heavy mental burden on the operator and requires a large number of key operations to obtain the desired word. For example, according to a conventional device (word processor WD-2400, etc.), as shown in FIG. 1 (ta), when "Usei" is entered in kana in the part to be converted, it is displayed in reverse video on the screen l. Next, when you operate the circle key, the first candidate is displayed, and the number of homophones is highlighted (Fig. 1(b)). Next, when the p key is operated, the second candidate is displayed, and the number of homophone blue meaning words is subtracted by 1 and displayed in reverse video (FIG. 1(C)). As the i key is further pressed, the number of homophones decreases by l (Figure 1(d) shows an example of displaying the fifth candidate). As described above, with this type of conventional device, the operation is difficult for the operator, as it is difficult to know how many times the next candidate instruction must be given before the desired word or phrase will appear. In addition, in the conventional second method, the target of conversion in the text being created and the display position of the converted word group are far apart, so the operator's line of sight has to move considerably to select the desired word. Moreover, it is necessary to select a desired word from a large number of homonyms that are all displayed, which poses problems such as placing a heavy burden on the operator. <Object of the invention> The object of the present invention is to provide a homophone selection method with excellent operability that eliminates the above-mentioned conventional problems.
In order to achieve this object, the present invention provides a means for displaying candidates for conversion results during kana-kanji conversion in a sentence being created displayed in a main window -L on a display screen, and Means for displaying the homophones iff of the conversion results all at once above the saf window set directly below the conversion results displayed in the middle sentence, and the homonyms iff on the →j-b window. If the synonym group is displayed at the bottom of the display screen, place it directly above it on the right side of the display screen [11, 11
The display control means 21□ shifts the display to the left when it protrudes from the screen. <Examples of the Invention> The present invention will be described in detail below with reference to Examples. Before going into specific explanations of the embodiments of the present invention, we will first explain the multi-window display technology that forms the background of the present invention. This technology allows two types of information to be viewed simultaneously on the same display. In this case, in order to overlap two or more types of screens at the same time,
This occurs when a necessary part of the content of one display is hidden by the display of the other. In order to avoid this, these windows can be moved anywhere on the display screen, or even deleted, according to the user's instructions. Next, a specific example for realizing the multi-window display function will be described. There are various means for realizing the multi-window display function, and the present invention is not limited thereto. FIG. 2 is a block diagram showing the principle of a multi-window display function in a so-called bitmap CRT. In FIG. 2, 10 is a central processing unit (CPU). 11 is a CRT controller (CRTC), 12 is a CR
Window controller (wp
c), 13 is a frame memory for several planes having memory elements corresponding dot by dot to the display on the CRT, 14.15
and 16 are frame memories for one screen each constituting the frame memory 13, and 17 is a display screen of a CRT. Also, the addresses (+)l, q+, rl -s+ ) and (
01, Vl, wl, Xl, ) are the addresses of the four corners of each window of the frame memories 15 and 16, respectively. The address information for each of them is held within the window controller (WDC) l2. Also address (1)2, q2, r21 S2)
and (u2.V2 2 + x2) are the above addresses, W (1)I-Ql-rllsl) and ('l+Vl+W
1', Xl), and each address information is also held in the window controller (WDC) 12. Furthermore, the above window controller (WDC) l
2 holds display priority information between each window, and in the example shown in FIG. 2, (uvwx). (pqrs) order, i.e. frame memory 16.15.1
The display priority order is maintained so that the images are displayed in the order of No. 4. In the above configuration, the CRT controller (c
Rrc) 11 is the window controller (wt>c) 1
In addition to the above-mentioned frame memory 15 and ], each window address information of 6 held by 2, each window address information and display priority information on the CR7 screen 17, each frame memory 14, 15 and 14 of frame memory 13 are 16 as appropriate, the stored contents are read out, and the contents are displayed on the CR7 screen 17 as shown in FIG. In addition, by appropriately rewriting each of the above information held by the window controller (WDC) 12 according to a command from the CPU10, the display on the CR7 screen 17 is changed to the one shown in FIG.
This can be easily changed as shown in (c). That is, by rewriting the window address on the frame memory 13, display scrolling within the window becomes possible. For example, window address (ul, Vl, Wl
, Xi) by one line (by increasing each y-coordinate of the window address by -1 line), the display within the window is scrolled as shown in FIG. 3(a). □ Also, by rewriting the window address on the CR1 screen 17-11, the window can be moved. For example, by subtracting the window address (12, V2, W2. I won't do it. Furthermore, change the window priority order, for example, change the window address priority order to (p', q, r, 5) (u, v,
By changing the order of w, x), the display can be changed as shown in Fig. 3 fc). By using multi-window display technology in this way, there is no need to rewrite the dot patterns in the frame memories 14 to 16 at all, even when making major changes to the screen display as shown in Figure 3 (al to tc+). , it is only necessary to rewrite a few parameters such as the window address held in the window controller (WDC) 12. Therefore, display changes such as those shown in FIG. This can be done in a short period of time, so the user will not feel any discomfort at all. Note that even with ordinary display devices that do not have a multi-window display function, the contents of the frame memory 14 can be rewritten each time. It is possible to change the display shown in Figure 3 ta+ to fcl.In this case, it will take more time to change the display than when using multi-window display technology.Next, the above-mentioned display An uninvented embodiment using technology will be described in detail. Fig. 4 is a block diagram showing an embodiment of the present invention applied to a Japanese word processor. In Fig. 4, 11 is a keyboard; , the keyboard ll has 1Ia4-, [11 keys,
6 12 is the central processing unit Jt, (CPU), 13 includes function keys such as the shift key.
14 is a controller for a display device such as a CRT (CRT).
-C), and this CRTC 14 has a built-in window controller (wDc) 141. 15 is CR'r
It is a frame memory with a storage capacity of several planes corresponding to 160 display screens and 1 dot each, and 17 and 18 display screens.
are frame memories for one screen that constitute the frame memo U 15, respectively, and the memory 17 is the frame memory for one screen that constitutes the frame memo U 15.
A portion of memory 18 is displayed on screen 1 in the main window of
6 corresponds to the subwindow 161 of the homophone. Further, 19 is a character generator that stores character patterns, and 20 is a so-called character code buffer memory.
2 each. Further, 23 and 24 are p register and q register, respectively, and the value of p register 23 indicates the number of homophone candidates displayed in the main window, and the value of q register 24 indicates the homophone candidate displayed in the main window. The display is configured to indicate whether or not a window is displayed as q-0 (no window displayed) or q-1 (window displayed). Next, the operation of the apparatus configured as described above will be explained with reference to the operation flow shown in FIG. In the above configuration, if you input kana information from the keyboard 11, the input kana characters will be processed by the CPU.
The character code is stored in the sofa 21 via l2, and the character pattern is stored in the character generator J.
The data is read out and written into the frame memory 17, and the content stored in the frame memory 17 is displayed in reverse on the display screen 16 [FIG. 6(a)]. At this time,
gidp=q= of p register 23 and q register 24
It remains o. Note that the window controller (WDC) 141 has a CP.
Display flk between windows for instructions of tJ12
Character information is stored in the current character code buffer 21, which is set to give priority to frame memory 18 or frame memory 17i, and "...of the device" is stored in the character code buffer 21.
Assuming that the kana-kanji conversion process up to this point has been completed,
The kana information "Kosei" input after that will be converted. Next, [m] is displayed on the keyboard 11 for the kanji conversion process
When the key is operated, the flow shown in Fig. 5 (step nl,
A conversion routine is drawn according to n2). This conversion routine (step n2) corresponds to the kana-kanji conversion module 13 shown in FIG.
j: It is configured to return one kanji character string (kana-kanji conversion result) depending on the kana character string given from the CPU 12 and the value of the p register 23. is searched to obtain the inverted character string "Kosei", and the kana character string is given to the conversion module 3 along with the information that p=o. Conversion module] 3 first converts p=I)+ in the conversion routine.
1 = 1, and the first candidate of the corresponding conversion result "
A new report will be returned to CPtJ, 12 with the value of p=l and the total number of homonyms in the conversion result as ``12''). The CPU 12 uses this information to write the value p=1) into the p register 23. Also, write the character string "Welfare" and the total number of homophones "12" in the character code /<・sofa 21, read out the character 7 NO turn from the character generator 19, and read out the character string "Kengo" and the total number of homophones "12" in the frame memory 171r, ! +
It's crowded. Therefore, in the step 113, the first candidate of the conversion result is displayed on the main window of the display screen 16 as shown in FIG. 6 (bl), and the number of homophones is displayed in reverse video. At this time, the value of q register 24 is q−Q
remains the same 1゜Here, when the BLACK key 3 on the keyboard 11 is operated, the flow shown in Figure 5 (nl□~rzy)
Therefore, a sub homophone window is set and a plurality of sub homophone windows are set in 611 [in the example, 5
) conversion candidates are displayed, and the next candidate kanji string is displayed in the main window. That is, when the mouth key 66 is operated in step n1□, it is first determined in step n+2 whether the number of candidates is 0.1, and in this case the number of candidates is "12".
) The process moves to step n13, and in step n13 it is determined whether or not the screen has changed, and in this case, based on the information of Q=O, it is determined that the screen has changed, and the step knob n1 is
4, and in step 1114 p=2.3.4
．． We obtained candidates corresponding to 5, gathered new wind data, and asked for the index "艷" at the top of the list. Without skipping step n15, the previous candidate is skipped, and a plurality of candidates (5) are displayed in a sub-window together with the first candidate obtained at the beginning (n16). Next, the process moves to step n17, executes the conventional next candidate routine, reads out the second candidate, and displays it on the main window.
(Preparations for displaying the second candidate homophone are completed. FIG. 7 is a flowchart showing details of the processing operation when the 4 key is operated, and steps n12 to n in FIG.
The processing input flow corresponding to step 17 is step n 11'l
"n120 and n130-n144G are shown. To explain the operation of creating a new window when operating the n key in Fig. 7, first step ypn
In step 113 (set register R in PU12 to 1 to clear the window, then write the first candidate in the Rth (first) position of the window (in this case, "welfare"). Next, R = R + 1 ntts+n117), increase p by 1 until the above condition is reached (p =9+1) Search the dictionary 13 for the next candidate kanji string, return it to the CPU 12, write the candidate in the Rth position of the window, and repeat steps n1.8 to n1□0 of executing R=R+1. Through such operations, the kanji string "composition" corresponding to P=2.3.4.5,
``Fairness,''``Rehabilitation,'' and ``Offensiveness'' are obtained. Then, the character codes of the first to fifth candidates obtained by this operation are 9
Write to the character code buffer 22 and write “
Write the character code of index 1゛1'' before "Composition". Next, read these character patterns from the character generator 19, and set the window address (pl-Qi+rt+Sl) in the area where the five homophone phrases are written. Next, while executing R=R-1, the previous candidate routine is executed until R=1, and a blank feed of the previous candidate is executed to return the contents of the p register 23 to the previous state. Return to the state of (p=1).Next, set the q register 24 to 1 and display the window address 'I)z, Q2, r2゜5□) on the display screen 16.
The display target positions of the upper sub-homophone window are also determined, and the five homophones are displayed on the window 161 (Fu33'). Next, the next candidate processing routine of step n144 is executed, and the second candidate kanji string "composition" corresponding to p=2 and rllJ
is written into the character code buffer 21, the pattern is read out from the character generator 19 and written into the τ frame memory l7', and the second candidate of the conversion result is displayed on the main window of the display screen 16 as shown in FIG. along with the number of homophones. Also, in the state of tc+ in Fig. 6, if you press the [gray ij lokey]
Steps nll"n13 and n1g, n16- in the figure
The next operation is executed according to the flow of n17. In other words, the display itself of the homophone window 161 does not change, but the display of the index ゛°-'' moves down one position, and the next candidate is displayed on the main window by the same operation as before (rt, y). Figure 8 shows step nll l n1ll I n112.
I n121-n123, n1411 n143
and n l 44' This is a detailed flow corresponding to the above operation, and first step n111 ・n112- "1
21 * Graduation of n122 and n123 are all “NO”
”, and the judgment of Steve Tubu n141 was “NO”.
", and in step n143, the code of the O) finger mlJ in the character code bar sofa 22 is shifted down by one line, and the pattern of the index "艷" is also written down by one line on the corresponding frame memory 18. Next, the next candidate routine (n
raa) to obtain the kanji string "Kei" corresponding to p=3 and the remaining number of homophones "lO". Then the fifth
When you display the window according to the flow in the figure, the index "-" will go down by one on the homophone display window, and the kanji string "fairness" corresponding to p==3 will be displayed on the main window.
and the number of remaining homophones "10" is displayed. If there is another next candidate key-in, the above operation is repeated. Therefore, if the next candidate key-in continues three times from the state of tc+ in FIG. 6, the main window will display "Offensive". "08" is displayed on the homophone subwindow, and the indicator "-" moves to the left side of "Offensive", resulting in the display screen shown in Figure 6 (dl). Also, the state of Figure 6 (C) ( When the key is pressed at p = 2), the next 5 candidate groups are displayed on the sub window according to the flow of steps ``20'' 19 + n 14 to n17 in Figure 5. FIG. 8 is a flow diagram showing details when the f and p keys are operated. In FIG. 8, steps n20'l and l12o3-
If the determinations in n2o4 are all "NO", the process moves to step n2D5i and performs the operation of subtracting the next candidate (n217.) until the value of p is evenly divisible by 5. Therefore, in this case, we get the kanji strings ``fairness'', ``1 rehabilitation'', and ``offensive'' that correspond to p=3.4.5, but we do nothing with these kanji strings and simply discard them. Next, set R=l to clear the sub window display information, and determine whether the value of p has reached the maximum value and whether the value of straddle R has reached 6 (n2oy, 1208). , the search routine for the next candidate until the above condition is reached, the candidate is written in the Rth position of the window, and the operation of R=R+l (n209-
n2+1) is repeated. With this operation, P=6
, 7. Kanji string corresponding to 8.9.10 "Karasho", 1 proofreading", "
Obtain ``posterity'', ``posterity'', and ``fixed star''. Then, the character codes of the 6th to 10th candidates obtained by this operation are written into the character code buffer 22i, and P-
The character code of index II and II is inserted before "correction" corresponding to 6. Next, λ1. These character patterns are read from the character generator 19 and inserted into the frame memory 18 in which the window address (p+, qzrzs+) is set. Next, while R=R-1 is executed until R=l, the previous candidate routine is executed and the previous candidate routine is executed. The empty feed of fli is executed by 1''j to return to the state of p=5 (021□~n2□4). Next, set Q=1 and update the 5 homophones in the sub window. Display on 1610 (n2+5)
n Next, execute the next candidate routine in step n216 and p=
The sixth candidate kanji string "Kosei" and "7" corresponding to 6 are stored in the character code buffer 21, and the pattern is read out from the character generator 19 and stored in the frame memory 17.
on the main window of the display screen 16.
Display the 6th candidate of the conversion result along with the number of remaining homonyms as shown in el.
f, By pressing the 1 key, the contents of the homophone window change by the same operation as above, and the 11th and 12th candidates are displayed, and the 11th candidate is displayed in the main window (
Figure 6(f)). Furthermore, press and hold the [Gray i Punishment Lokey, or d, q
By pressing the key, the display of the homophone group becomes apparent, and the display becomes as shown in Figure 6 (gl).In addition, in any of the states shown in Figure 6 (bl~[gl), the kana of the next clause is manually input from the keyboard 11. The conversion target candidate is determined by Sukokichi, and the process moves to the next step.For example, in the state shown in FIG. 6 tel, if a kana character is input from the keyboard ll, the flow shown in FIG. First, clear the p and q registers 23 and 24 to p-O, q-○, erase the window, and write ', on the character code buffer 21.
-jf Delete 1xlH and insert the new character code "yo". Next, save that character pattern to frame memory 1.
Write to 7. By performing the same operation for the kana characters ``u'' and ``so'', a surface screen as shown in FIG. 6(h) is obtained. Here, by pressing the [ED key ~, the kana-kanji conversion for the kana input "yoso" is executed, and a screen displaying the first candidate similar to that shown in Fig. 6 (1)) is displayed. You can choose to display the accompanying edict. In addition, in FIG. 7G, steps 0123 to 111□9
This flowchart shows the processing flow for updating the display of the sub window when the m key is operated in the screen state shown in Fig. 6 (td), and the processing flow is described in Fig. 8 above. Processing steps n 207 to n2
Since it is the same as No. 11, its explanation will be omitted. Furthermore, Fig. 9 shows the specific processing flow when operating the -Ii lokey, and Fig. 1Q shows the
This shows the specific processing flow when you press the ``1''' key, and each operation moves the displayed candidates back to the previous one.
The detailed operation is similar to that shown in FIGS. 7 and 8, except that the directions of increase/decrease in the P register and increase/decrease in the R register are reversed, so a detailed explanation will be omitted. According to the embodiment described above, as shown in FIG. are doing. Therefore, it is difficult to select a desired homophone while knowing in advance what homophone will be output by the next N key operation. Also d
, [

Since the homophone display on the multi-window is updated all at once using the [Possible i] key, fewer key operations are required to obtain the correct answer candidate. Furthermore, when compared with the conventional method, there is no difference between the two when the first or second candidate is correct. When one of the third to fifth candidates is correct, there is no difference in the number of key operations between them, but when it turns out that the second candidate is not the correct answer, the homophone group on the sub multi-window is If you look at it, you can confirm that the third to fifth candidates are correct, and the operator can operate the L Country Lokey with peace of mind. Also, when any of the 6th to 1st candidates is correct, when it becomes clear that the 2nd candidate is not the correct answer, if you look at the homophone group on the multi-window, you will see that the 3rd to 5th candidates are also not correct. You can check immediately []〒E,
By operating the J key, the display of the 6th to 10th candidates can be obtained, and the number of key operations required to obtain the correct answer can be reduced compared to the conventional method. Furthermore, if the 11th or 12th candidate is the correct answer, it is possible to further reduce the number of key operations. The above example shows a case where the display position of the homophone window is directly below the display position of one homophone of the conversion result displayed on the main window, but the present invention further provides that the homophone window is It has a function to display it directly above when it protrudes from the bottom of the display screen, and to shift it appropriately to the left when it protrudes to the right. FIG. 11 is a processing flow for determining the display position of the homophone window to realize the above function. As shown in FIG. 12, the display capacity of the display screen 16 is 40 characters horizontally and 20 characters vertically, and the top position of the homophone display of the conversion result on the main window is (x,y).
Set the position of address P2 of the sub window to (x, ,
y ), then CI) tJ 12 must first grasp the starting position (X, Y) of the conversion target. Step n5 in Figure 11
o1') First, calculate the height H and width W of the subwindow (n602). This calculation is performed, for example, using the window address (pl + Ql, rl
, sl), the value of H is calculated by subtracting the Y coordinate of address prQ) from the X coordinate of address r1, and the value of W is calculated by subtracting the X coordinate of address p1 from the X coordinate of address q1. Calculate by Next, proceeding to step n503, it is determined whether Y+H>20 or not, and if Y10>20, the calculation of y - Y -1-1 is executed in step nBO4, and τT address 2
The X coordinate position of the address p2 is detected, and if y+H>20, the process moves to step n5os, and the calculation of Y=Y-H is executed to calculate the X coordinate position of the address p2. That is, step n' 5
In step n6051, if the homophone window is displayed or the normal display is performed, it is determined whether or not it protrudes to the bottom of the display screen. Window address 1 so that it is directly above the window conversion result display
) Determine the X coordinate of 2. Next, it moves to step n506 and determines whether or not X 10w>42, and if
Execute the calculation of x = 4.1-w to calculate the X coordinate position of address 1) 2, and if x+W>42, n5oy
Step 11508 moves to step 11508, executes the calculation x=X-1, calculates the X coordinate position of address p2, and determines whether or not X=1. If i, move to step n5oglc and address p2f7)X
Let the coordinates be x=1. That is, in step n506, it is determined whether the sub window's display power 5 will protrude to the right side of the display screen when normal display is performed, and if it is determined that it protrudes, the sub window The X coordinate of the window address p2 is determined so that the two displays are appropriately shifted to the left of the display of the conversion results in the main window. Next, the process moves to step n511, where (x, y) is set at the upper left position, and a window with a height of H1 and a width of W is displayed. That is, the window display technique described above is performed by determining the sub window address as 5 p2 = (x, y) q2 = (x + W-y) r2 = (x, y + H) s2 = (x + W, y-1-H) It will be displayed according to the following. Then, as a result of the processing operations described above, the sub window is displayed on the lower side of the conversion target in the main window as shown in Figure 13 (a+), and the left edge of the conversion target and the same sound on the sub window The left side of the word display matches measles (Ste Tsubu n504-n5011 + n5
11). Furthermore, when the conversion target is at the bottom of the screen, the sub window is displayed above it as shown in FIG. Even when the object is bent to the left end, the tip of the object to be converted is the reference point for the sub window display, but as shown in Figure 13 tc+, the sub window display protrudes from the right edge of the display screen. It will be displayed shifted to the left (step n5o4+
Based on the determination of n5lo -n,su. ). Note that the type of each key in the above embodiment is not limited to this, and may be used in combination with other keys, or may be arbitrarily changed such as providing a dedicated key. <Effects of the Invention> As described above, according to the present invention, the display of the homophone group on the sub window can be displayed at the bottom of the display screen (if it protrudes beyond the can, it will be displayed directly above it, and it will also protrude to the right side of the display screen 0). Sometimes it is shifted to the left to display it, so
The homophone group is optimally displayed in a sub-window near the display of the sentence being created, and the movement of the operator's line of sight when selecting a desired homophone can be minimized. At the same time, it becomes possible to perform the selection operation extremely efficiently.

[Brief explanation of the drawing]

Figure 1 is a diagram showing a display example of the homophone selection operation during conventional kana-kanji conversion, Figure 2 is a functional diagram of the multi-window display that is the background of the present invention, and Figure 3 is the multi-window display. 4 is a program diagram showing an embodiment of the present invention applied to a Japanese word processor, and FIG. 5 is a diagram showing the operation flow of an embodiment of the present invention. , FIG. 6 is a diagram showing an example of display when selecting, selecting, and selecting a homophone group displayed according to the present invention, FIG. 7 is a diagram showing a detailed operation flow corresponding to the next candidate key operation, and FIG. 5
HIFT, a diagram showing the detailed operation flow corresponding to the next candidate key operation, FIG. 9 is a diagram showing the detailed operation flow corresponding to the operation of the previous candidate key, and FIG. 10 is 5HIFT, previous candidate? fli
Figure 11 is a diagram showing the detailed operation flow corresponding to key operations, Figure 11 is a diagram showing the operation flow for determining the display position of the strip homophone window, Figure +2 is a diagram showing the position of the lever on the display screen, Figure 13
The figure is a diagram showing a display state of homophone groups displayed according to the present invention. 11...Keyboard, 12...Central processing unit (cp
u) 13...Kana-Kanji conversion dictionary 1.14...CRT controller, 16...Display screen, 17.18...Frame memory, 21. 22...Character code buffer, 161...Sub window display area 1, agent
Patent attorney Aihiko Fukushi (and 2 others) tσノ(b) 1st (C) (d)

Claims (1)

[Scope of Claims] 1. Means for displaying one candidate of the conversion result during kana-kanji conversion in the text being created displayed on the main window of the display screen; means for collectively displaying the homophone group of the conversion result on a sub-window set immediately below the displayed conversion result; and a means for displaying the homophone group of the conversion result on the bottom of the display screen. A homophone display selection method characterized by comprising a display control means for displaying a homophone immediately above the display screen when the display screen extends, and for shifting the display to the left when the display screen protrudes to the right of the display screen.