JP5482236B2 - Program and information processing apparatus - Google Patents

Description

  The present invention relates to a program and an information processing apparatus.

  A technique for supporting text input is known. For example, Patent Document 1 discloses a technique for predicting a character string following a character string including a kanji or hiragana input by a user. In the technique described in Patent Document 1, a plurality of character strings composed of hiragana, kanji, or mixed kanji are stored in a dictionary, and a character string extracted from the above dictionary using the input character string as a key is stored. And output as a predicted character string following the input character string.

  In Patent Document 2, on a pen input computer, one or more candidate words retrieved from a dictionary storage unit that stores a plurality of words and a plurality of sentence examples are displayed on a display unit based on input and / or already determined character strings. A technique for inputting text by allowing a user to select a desired word from the displayed candidate words is disclosed.

  Patent Document 3 discloses a technique for accepting input of a sentence including an ellipsis and complementing a character string corresponding to the ellipsis. In the technique described in Patent Document 3, a word dictionary that stores words and the frequency of use of the words is extracted, word candidates omitted from characters before and after the ellipsis in the input sentence are extracted, Based on the transition dictionary storing the transition information and the transition probability between words, one word is determined from the extracted candidate words.

  There are also technologies that support the proofreading of entered text. For example, in the technique described in Patent Document 4, a sentence to be processed is analyzed in a unit of phrase, and a conventional use composed of a plurality of elements that are conventionally connected in a sentence from each phrase. By detecting the constituent elements of the phrase, and further rearranging (rearranging) the detected constituent elements, it is possible to assist in proofreading the sentence more easily.

JP-A-8-255158 Japanese Patent Laid-Open No. 10-154033 JP 2000-330984 A Japanese Patent Laid-Open No. 7-234871

  By the way, when correcting the input text, the user may desire to insert a character string or may want to replace a part of the character string in the text.

  It is an object of the present invention to provide a program and an information processing apparatus that support both correction of text accompanying insertion of a character string and correction of text accompanying replacement of a character string.

The invention according to claim 1 refers to a reception step of receiving a specified position that is a position in a target character string that is a character string to be processed, and a dictionary storage unit that stores a plurality of sentences, and the specification in the target character string A character string included up to the position as a search key, and in a sentence including the search key character string among the plurality of sentences stored in the dictionary storage means, a character string following the search key character string, Guessing to infer as a candidate character string that is a candidate for a character string following the character string up to the specified position in the target character string, and performing a morphological analysis on the character string after the specified position in the target character string an analysis step, based on a result of the morphological analysis, the string after the specified position in the target string, the element string is a character string corresponding to the components of the sentence The character string in the target character string after the element character string is added to the candidate character string based on a predetermined evaluation rule for each of the pair of the candidate character string and the element character string. An evaluation step for evaluating an expression followed by a sequence, obtaining an evaluation value representing the validity of the expression as a natural language, and an output step for outputting the expression having the evaluation value equal to or greater than a preset threshold value It is a program for making it run.

  According to a second aspect of the present invention, in the first aspect of the invention, the characters after the designated position in the target character string in response to a user input selecting one of the expressions output in the output step. A generation step of generating a character string in which a string is replaced with an expression selected by the input is further executed by the computer.

  The invention according to claim 3 is the invention according to claim 1 or 2, wherein, in the output step, the evaluation value is higher when there are a plurality of expressions whose evaluation value is equal to or greater than the threshold. Display control information for controlling the display unit to display the display with priority is output.

  The invention according to claim 4 is the invention according to any one of claims 1 to 3, wherein the evaluation value is not obtained for the same expression as the expression included in the target character string.

  The invention according to claim 5 is the invention according to any one of claims 1 to 4, wherein, in the output step, the candidate character string and the target character in the expression in which the evaluation value is not less than the threshold value. Display control information for controlling the display unit to display the character string in the column in a different manner is output.

  The invention according to claim 6 is the invention according to any one of claims 1 to 5, wherein, in the output step, in the target character string, the designated position of the target character string and the evaluation value are Display control information for controlling the element character string arranged between the element character string included in the expression that is equal to or greater than the threshold to be displayed on the display unit together with the expression is output.

The invention according to claim 7 refers to a receiving unit that receives a specified position that is a position in a target character string that is a character string to be processed, and a dictionary storage unit that stores a plurality of sentences, and the specification in the target character string A character string included up to the position as a search key, and in a sentence including the search key character string among the plurality of sentences stored in the dictionary storage means, a character string following the search key character string, Guessing means for estimating a candidate character string that is a candidate for a character string following the character string up to the specified position in the target character string, and performing a morphological analysis on the character string after the specified position in the target character string analyzing means, based on a result of the morphological analysis, decompose decompose string after the specified position in the target string, the element string is a character string corresponding to the components of the sentence For each pair of the candidate character string and the element character string, an expression in which the character string in the target character string subsequent to the element character string is followed by the candidate character string according to a predetermined evaluation rule. Evaluation means for evaluating and obtaining an evaluation value representing the validity of the expression as a natural language; and output means for outputting the expression whose evaluation value is equal to or greater than a preset threshold value. Information processing apparatus.

  According to the invention which concerns on Claim 1 or 7, both correction of the text accompanying insertion of a character string and correction of the text accompanying substitution of a character string can be supported.

  According to the invention which concerns on Claim 2, a text can be corrected by the expression which the user selected.

  According to the third aspect of the present invention, it is possible to preferentially present an expression having high validity as a natural language to the user.

  According to the invention which concerns on Claim 4, the process which calculates | requires an evaluation value about the same expression as the expression in an object character string can be abbreviate | omitted.

  According to the fifth aspect of the present invention, the candidate character string estimated as the character string following the designated position and the character string originally included in the target character string can be distinguished and presented to the user.

  According to the invention which concerns on Claim 6, the character string in the original object character string substituted by the candidate character string in the said expression can be shown to a user with the expression produced | generated using the candidate character string.

It is a block diagram which shows the example of the outline of an internal structure of information processing apparatus. It is a figure which shows the example of a display screen. It is a figure which shows the example which decomposed | disassembled the text into the morpheme. It is a figure which shows the example of the expression which connected the candidate of the subsequent character string of a designated position. It is a figure which shows the example of the procedure of the process which information processing apparatus performs. It is a figure which shows the example of the procedure of a connection expression evaluation process. It is a figure which shows the other example of the expression which connected the candidate of the subsequent character string of a designated position. It is a figure which shows the example of a part of result of a connection expression evaluation process. It is a figure which shows the example of a part of result of a connection expression evaluation process. It is a figure which shows the example of a part of result of a connection expression evaluation process. It is a figure which shows the example of a part of result of a connection expression evaluation process. It is a figure which shows the example of the aspect of a display. It is a figure which shows the other example of the aspect of a display. It is a figure which shows the further another example of the aspect of a display. It is a figure which shows the example of the hardware constitutions of a computer.

  FIG. 1 shows an outline of an internal configuration of an information processing apparatus according to an example of an embodiment of the present invention. The information processing apparatus 10 in the example of FIG. 1 includes a document storage unit 100, a reference data storage unit 110, an input reception unit 120, a target text acquisition unit 130, a subsequent character string estimation unit 140, a text decomposition unit 150, and a connected expression evaluation unit 160. , An output processing unit 170, and a modified text generation unit 180.

  The document storage unit 100 stores an electronic document. Processing of each unit of the information processing apparatus 10 to be described later is performed on a text (character string) in the electronic document stored in the document storage unit 100 as a processing target.

  The reference data storage unit 110 stores information referred to in processing performed by the information processing apparatus 10. The reference data storage unit 110 includes an analysis dictionary 112, a sentence example dictionary 114, and evaluation data 116. Details of the analysis dictionary 112, the sentence example dictionary 114, and the evaluation data 116 will be described later.

  The input receiving unit 120 receives user input via an input device such as a mouse and a keyboard. The input reception unit 120 passes the received input information to the target text acquisition unit 130, the subsequent character string estimation unit 140, or the corrected text generation unit 180 according to the content.

  The target text acquisition unit 130 acquires the text to be processed from the electronic document stored in the document storage unit 100. The target text acquisition unit 130 acquires, for example, character strings before and after a position designated by the user in the electronic document as a processing target. For example, referring to FIG. 2, in the display screen 200 in which the electronic document stored in the document storage unit 100 is displayed on the display device (not shown), the user moves the cursor to the position indicated by the arrow C using the input device. To do. In this example, the target text acquisition unit 130 acquires information indicating the position of the cursor via the input reception unit 120, and regards the cursor position as a specified position by the user, and acquires character strings before and after the specified position. . The character string acquired at this time may be, for example, a character string constituting one sentence including a specified position. In the example of FIG. 2, a sentence including the designated position “A high absorption area is seen in the right lung” is acquired. For example, a sentence containing the specified position is determined by checking the forward and backward character strings one by one from the specified position and finding a symbol (punctuation mark, exclamation mark, question mark, etc.) that indicates a sentence break. It is obtained by acquiring a character string between the symbols.

  In the description of the example of the present embodiment, the “front and back” character string at the designated position means the “left and right” character string at the designated position in a horizontally written document, and designated in a vertically written document. It means the character string "up and down" of the position.

  The subsequent character string estimation unit 140 estimates a character string candidate that follows the specified position in the text to be processed according to a predetermined estimation rule. In the case of the target text in the example of FIG. 2, the subsequent character string estimation unit 140 estimates a character string candidate that follows the character string “in the right lung” up to the designated position. In the example of the present embodiment, the subsequent character string estimation unit 140 estimates a candidate for the subsequent character string at the specified position with reference to the sentence example dictionary 114 provided in the reference data storage unit 110. The example sentence dictionary 114 is a dictionary that stores a plurality of sentences included in an electronic document stored in the document storage unit 100. The sentence example dictionary 114 may further store, for each sentence, grammatical roles such as word breaks in the sentence and part of speech of each word. The subsequent character string estimation unit 140 searches, for example, the sentence example dictionary 114 using the character string up to the specified position of the text to be processed as a search key, and selects a word or phrase following the character string in a sentence including the character string of the search key. Identify. Then, a specified word or phrase having an appearance frequency equal to or higher than a preset threshold is set as a subsequent character string candidate at the designated position. The candidates obtained in this way may be ranked in descending order of appearance frequency. Alternatively, for example, a score based on the appearance frequency may be determined for each candidate. The method by which the subsequent character string estimation unit 140 estimates the subsequent character string at the specified position is not limited to the above-described example, and various other conventionally known methods may be used.

  The text decomposition unit 150 analyzes the text to be processed acquired by the target text acquisition unit 130 according to a predetermined analysis rule, and decomposes each element constituting the sentence. In the example of the present embodiment, the text decomposition unit 150 refers to the analysis dictionary 112 provided in the reference data storage unit 110, and decomposes the text to be processed into morphemes using a morphological analysis method known in the field of natural language processing. To do. The analysis dictionary 112 of this example is a dictionary that stores a word and information indicating a grammatical role of the word in association with each other and stores Japanese grammar rules. In the case of the processing target text in the example of FIG. 2, the text decomposition unit 150 refers to the analysis dictionary 112 and exemplifies a sentence “a high absorption area is seen in the right lung” as illustrated in FIG. 3. Disassemble each word. The text decomposition unit 150 may also analyze and decompose the subsequent character string candidates estimated by the subsequent character string estimation unit 140. For example, when the subsequent character string estimation unit 140 estimates “significant” as a candidate for the subsequent character string at the specified position in the example of FIG. 2, the text decomposition unit 150 performs a morphological analysis on the character string, Disassemble for each word “Ming” and “N”.

  The connected expression evaluation unit 160 includes each of the subsequent character string candidates estimated by the subsequent character string estimation unit 140 and each element obtained by the text decomposition unit 150 decomposing the character string after the specified position in the text to be processed. For a set, the validity of the expression as a natural language in which a character string in a text to be processed after the element is connected to the candidate is evaluated. In other words, the likelihood of each expression is evaluated.

  Hereinafter, a specific example of the expression to be evaluated by the connected expression evaluation unit 160 will be described. Consider the case where the text to be processed in the example of FIG. 2 is decomposed as in the example of FIG. 3 and “significant” is estimated as a candidate for the subsequent character string of the character string “in the right lung” up to the specified position. In this example, the connected expression evaluation unit 160 evaluates “high absorption”, “range”, “ga”, and “seen” among the words (elements) after the specified position in the text to be processed. . Referring to FIG. 4, the character string “right lung” up to the designated position is followed by a candidate for the succeeding character string “significant”, and after “significant”, each word to be evaluated after the designated position The validity of each expression that concatenates the character strings is evaluated. In the example of FIG. 4, the expression “connected to each broken arrow a, b, c, d” “A remarkable high absorption area is seen in the right lung.”, “A remarkable area is seen in the right lung. . ”,“ A prominent look is seen in the right lung ”, and“ A prominent look is seen in the right lung ”are obtained.

  As can be seen from the above description with reference to FIG. 4, in the example of this embodiment, in the connected expression evaluation unit 160, an expression in which the candidate for the subsequent character string is inserted at the designated position of the processing target text (broken arrow a in FIG. Expression) (see broken line arrows b, c, and d in FIG. 4) in which a part of the character string in the processing target text is replaced with a candidate for the subsequent character string is generated. Sex is evaluated.

  Referring again to FIG. 1, the connected expression evaluation unit 160 uses the evaluation data 116 included in the reference data storage unit 110 to obtain an evaluation value for each expression according to a predetermined evaluation rule. The evaluation data 116 is data necessary for obtaining the evaluation value of each expression, and is generated in advance and stored in the reference data storage unit 110. The contents of the evaluation data 116 differ depending on the evaluation rule used by the connected expression evaluation unit 160. As the evaluation rule, a method conventionally used in the field of natural language processing when evaluating the validity of an expression as a natural language may be used. As an example, a probabilistic language model such as an N-gram model may be configured in advance, and an estimated value of the appearance probability of each expression may be obtained in accordance with this probabilistic language model and used as an evaluation value for each expression. In this example, the evaluation data 116 includes information defining a probabilistic language model. The information that defines the probabilistic language model includes, for example, the appearance frequency of each word and the frequency with which another word appears after a certain word. In another example of the evaluation rule, a parsing technique for analyzing the syntax of a character string according to a certain grammar may be used. When the parsing technique is used, the evaluation data 116 includes information representing the grammar rules of the language (for example, Japanese) of the text to be processed and a dictionary of words in the language. The linked expression evaluation unit 160 performs syntax analysis on each expression to be evaluated with reference to the grammar rules and the dictionary in the evaluation data 116, and determines an evaluation value of each expression based on the result of the syntax analysis. For example, a value indicating that the expression is valid as a natural language expression if the parsing is successful, and a value indicating that the expression is not valid as a natural language expression if the parsing fails, are used as the evaluation value of the expression.

  The output processing unit 170 performs a process of outputting a processing result in the information processing apparatus 10. For example, the output processing unit 170 selects an expression to be output according to the evaluation value of each expression obtained by the connected expression evaluation unit 160, and displays the selected expression on a display device (not shown). At this time, the output processing unit 170 may display the selected expressions in descending order of evaluation values, that is, in order of high relevance as a natural language. If the subsequent character string estimation unit 140 estimates a plurality of subsequent character string candidates and ranks these candidates or assigns a score to each candidate, the output processing unit 170 The expression to be output may be selected by further considering the candidate rank or score. A specific example of the display mode will be described later.

  The modified text generation unit 180 selects the user via the input reception unit 120 when the user selects one expression from the expressions output by the output processing unit 170 based on the evaluation result by the connected expression evaluation unit 160. The specified expression is specified, and processing for replacing the text to be processed with the selected expression is performed. By the processing of the corrected text generation unit 180, the processing target text is corrected to the selected expression in the electronic document including the processing target text.

  Hereinafter, an example of processing of the information processing apparatus 10 will be described. FIG. 5 is a flowchart illustrating an example of a procedure of processing performed by the information processing apparatus 10. The information processing apparatus 10 starts the processing of the procedure in the example of FIG. 5 when, for example, the user inputs to specify a specific position in the electronic document.

  First, the input receiving unit 120 acquires a designated position in the electronic document designated by the user using the input device (step S10).

  The target text acquisition unit 130 that has acquired the specified position from the input receiving unit 120 acquires the character string before and after the specified position and sets it as the processing target text (step S12). In the example of the present embodiment, the target text acquisition unit 130 acquires a character string constituting one sentence including the designated position in step S12. For example, a sentence including the specified position can be obtained by acquiring a character string between a specified position and a delimiter (a punctuation mark, an exclamation mark, a question mark, etc.) of a sentence that first appears before and after the specified position.

  The text decomposing unit 150 analyzes the processing target text acquired in step S12 and decomposes it into elements constituting the sentence (step S14). In the example of the present embodiment, the text decomposition unit 150 performs morphological analysis with reference to the analysis dictionary 112 in step S14, and decomposes the processing target text for each word.

  The subsequent character string estimation unit 140 that has acquired the specified position from the input receiving unit 120 estimates a character string candidate (hereinafter also simply referred to as “subsequent candidate”) following the character string up to the specified position (step S16). In the example of this embodiment, in step S14, the subsequent character string estimation unit 140 first searches the example sentence dictionary 114 using the character string of the word or phrase immediately before the specified position as a search key. Then, in a sentence in the sentence example dictionary 114 including the search key character string, the word or phrase following the search key character string is specified, and the appearance frequency of the specified word or phrase is equal to or greater than a preset threshold. The thing is a subsequent candidate. The character string of the search key may be determined with reference to the result of the decomposition process in step S14 by the text decomposition unit 150. For example, the part-of-speech of each word is examined in order from the word immediately before the specified position, and the first independent word (a noun, verb, or other word that can form a single phrase), and between the independent word and the specified position A character string including the positioned word may be used as a search key. Further, when a plurality of subsequent candidates are obtained, the subsequent character string estimation unit 140 may rank the plurality of subsequent candidates according to, for example, the appearance frequency of each subsequent candidate. Or you may determine the score of each subsequent candidate based on the appearance frequency of each subsequent candidate. The score of each subsequent candidate is a value representing the high possibility that the subsequent candidate follows the character string up to the designated position. Note that the subsequent character string estimation unit 140 may request the text decomposition unit 150 to decompose the obtained subsequent candidates into words.

  The connected expression evaluation unit 160 selects one of the subsequent candidates estimated in step S16 (step S18), and performs a connected expression evaluation process on the selected subsequent candidate (step S20). FIG. 6 shows an example of a detailed procedure of the connected expression evaluation process in step S20 of FIG. When step S20 of FIG. 5 is started, the processing of the procedure of the example of FIG. 6 is started.

  With reference to FIG. 6, the connected expression evaluation unit 160 first identifies an element to be evaluated among elements after the designated position in the processing target text (step S <b> 200). In the example of the present embodiment, after a specified position, a word constituting one clause and a word immediately after this clause are evaluated. In the case of the decomposed processing target text illustrated in FIG. 3, the words (“high absorption”, “range”, “ga”) included in the phrase “high absorption range” after the specified position and immediately after the phrase The word “seen” is the object of evaluation. In an example of another embodiment, a preset number of elements may be evaluated after the designated position. In still another example of the embodiment, all elements after the designated position may be evaluated.

  Next, one element to be evaluated is selected (step S202), and an expression in which a character string after the selected element is linked to a subsequent candidate is generated (step S204). In the case of the above example with reference to FIG. 3, if “high absorption”, which is one of the words to be evaluated, is selected in step S202, and the subsequent candidate is “significant”, then the connected expression The evaluation unit 160 generates an expression “a remarkable high absorption region is seen in the right lung” by concatenating a character string after “high absorption” to “significant” (see the broken arrow a in FIG. 4). To do.

  After step S204, the linked expression evaluation unit 160 determines whether or not the generated expression is the same as the original character string expression (step S206).

  For example, for the text to be processed in the example of FIG. 3, when the subsequent candidate is “high absorption” and the word “high absorption” is selected in step S202, the expression generated in step S204 as shown in FIG. “A high-absorption area is seen in the right lung” is the same as the original text to be processed. In such an example, YES is determined in step S206, the process returns to step S202, one of the unprocessed evaluation target elements is selected, and the processes in and after step S204 are repeated.

  On the other hand, the generated expression is different from the expression of the original processing target text as in the case of the expression in the above example with reference to FIGS. 3 and 4, in which a marked high absorption area is seen in the right lung. In this case, the connected expression evaluation unit 160 obtains an evaluation value representing the validity of the expression generated in step S204 as a natural language (step S208). In step S208, the linked expression evaluation unit 160 refers to the evaluation data 116 in the reference data storage unit 110, and obtains the evaluation value of the expression generated in step S204 according to a predetermined evaluation rule. In the example using the above-mentioned stochastic language model, when evaluating the expression illustrated in FIG. 4 “a remarkable high absorption region is seen in the right lung”, for example, using the evaluation data 116, the subsequent candidate “ What is necessary is just to obtain | require the probability that the character string "high absorption area" will appear after "bright", and let it be an evaluation value. Further, in the case of the above-described parsing example, parsing is performed on the expression “a significant high absorption area is seen in the right lung” using the evaluation data 116 including grammatical rules and a dictionary. The evaluation value of the expression may be obtained according to the success or failure of the syntax analysis.

  After step S208, if all of the evaluation target elements specified in step S200 have been processed (YES in step S210), the linked expression evaluation process in the procedure of the example of FIG. 6 ends. If not all have been processed (NO in step S210), the process returns to step S202 to select an unprocessed element to be evaluated, and the processes in and after step S204 are repeated.

  Referring to FIG. 5 again, until the connected expression evaluation process (step S20, FIG. 6) is performed for all the subsequent candidates estimated in step S16 (until YES is determined in step S22), selection of subsequent candidates (step S18) and the connected expression evaluation process (step S20) are repeated.

  When the connected expression evaluation process is completed for all subsequent candidates (YES in step S22), the output processing unit 170 determines an expression to be output (step S24). In step S <b> 24, the output processing unit 170 determines the output target expression using the evaluation value obtained in the connected expression evaluation process for the combination of each subsequent candidate and each evaluation target element of the connected expression evaluation unit 160. To do. In addition, when the plurality of subsequent candidates estimated in step S16 are ranked or when the score of each subsequent candidate is determined, the output processing unit 170 further considers the rank or score, The expression may be determined.

  Hereinafter, a specific example of step S24 will be described with reference to FIGS. 8A to 8D. FIGS. 8A to 8D show the connected expression evaluation process for each subsequent candidate when four prominent candidates, “significant”, “high absorption”, “low absorption”, and “tumor” are estimated in step S16. The example of the evaluation value calculated | required by (step S20, FIG. 6) is shown. FIG. 8A is an example of an evaluation value of an expression including the subsequent candidate “striking”, and shows an evaluation value corresponding to each broken-line arrow in the same diagram as FIG. FIG. 8B shows an example of an evaluation value of an expression including the subsequent candidate “high absorption”. FIG. 8B shows the evaluation value of each expression in which the subsequent candidate “high absorption” is connected with the character strings after the words “high absorption”, “ga”, and “seen” in the text to be processed. Referring to FIG. 8B, the evaluation value of the expression in which the word “region” in the text to be processed is connected to the subsequent candidate “high absorption” is not calculated. This is because it is the same as the original processing target text expression, so that a YES determination is made in step S206 of the linked expression evaluation process in the example of FIG. 6, and evaluation value calculation (step S208) is not performed. FIG. 8C shows the evaluation value of each expression in which the subsequent candidate “low absorption” is concatenated with the character strings after the words “high absorption”, “range”, “ga”, and “seen” in the text to be processed. Similarly, FIG. 8D shows the evaluation value of each expression in which the subsequent candidate “tumor” is connected to the character string after each word in the text to be processed.

  When the rank and score of the subsequent candidates are not considered in determining the expression to be output, for example, the output processing unit 170 selects an expression whose evaluation value is equal to or greater than a preset threshold as an output object. As an example, when this threshold value is 0.4, for the subsequent candidate “significant”, the expression of the evaluation value 0.6 “a remarkable high absorption range is seen in the right lung” is selected ( FIG. 8A). The subsequent candidate “high absorption” is not selected as an output target because there is no expression with an evaluation value of 0.4 or more (FIG. 8B). For the subsequent candidate “low absorption”, an expression of an evaluation value of 0.6 “a marked low absorption region is seen in the right lung” is selected (FIG. 8C). For the subsequent candidate “tumor”, the expression “a tumor is found in the right lung” with an evaluation value of 0.5 is selected (FIG. 8D).

  When considering the rank or score of the subsequent candidate in the determination of the expression to be output, for example, the output processing unit 170 has a preset threshold value obtained by weighting the evaluation value of each expression according to the rank or score of the subsequent candidate. The expression to be output may be determined according to whether or not this is the case.

  When the output target expression is determined, the output processing unit 170 outputs the determined output target expression (step S26). In the example of this embodiment, in step S26, the expression to be output is displayed on a display device (not shown). For example, the expression of the output target is displayed together with the display screen on which the electronic document including the processing target text is displayed. At this time, for example, the expressions to be output may be displayed in the form of a list in which the evaluation values are arranged in descending order. Further, for example, the expressions to be output may be displayed side by side in descending order of evaluation values weighted according to the rank or score of the subsequent candidates. The output processing unit 170 realizes the process of step S26 by, for example, generating display control information that instructs a display mode as in each example described later and outputting the display control information to the display device.

  9 to 11 show various examples of modes for displaying the expression of the output target. FIGS. 9 to 11 show the expression “a remarkable high absorption region is seen in the right lung” (FIG. 8A) and “right lung” in the specific example of step S24 described with reference to FIGS. 8A to 8D. Is a specific example of display in step S26 when “A tumor is seen in the right lung” (FIG. 8C) is selected as an output target.

  FIG. 9 and FIG. 10 enumerate the expressions of the output targets, and the subsequent candidate character strings, the character strings in the original processing target text, and the character strings replaced by the subsequent candidates in the processing target text are different from each other. An example of display is shown. In the example of FIG. 9, the character string of the subsequent candidate is displayed in bold, the character string replaced by the subsequent candidate is displayed in parentheses after the subsequent candidate, and the character string in the other processing target text is displayed as the subsequent candidate character. Display with a font that is thinner than the column. In the example of FIG. 10, a line is drawn and displayed on the character string replaced by the subsequent candidate. In the example of FIG. 10, the display mode of other character strings replaced by subsequent candidates is the same as the example of FIG. The display screen in the example of FIG. 9 or FIG. 10 may be displayed on the display device together with the display screen (see FIG. 2) on which the electronic document including the original text to be processed is displayed. For example, the display screen of the example of FIG. 9 or 10 may be displayed near the display position of the text to be processed on the display screen of the electronic document of the example of FIG.

  In the example of FIG. 11, the display position of the common portion between the character string of the text to be processed and the character string of the expression to be output is displayed together. In the example of FIG. 11, the character string common to all of the processing target text and the output target expression is “seen”. For each expression of the text to be processed and the output target, the character string “I can see” is displayed so as to be in substantially the same position in the horizontal direction of the display screen. Further, in the expression to be output, the display of the character string before the succeeding candidate (that is, the character string up to the designated position) and the character string after the above-mentioned common part “seen” is omitted. In the expression to be output, each subsequent candidate character string is displayed in a thicker font than the other character strings. In the example of FIG. 11, another character string of the subsequent candidate in the output target expression may be displayed in a lighter (or lighter) color than the character string of the subsequent candidate and the character string of the processing target text.

  You may display in the other aspect of the example demonstrated with reference to FIGS. For example, the subsequent candidate character string, the character string in the processing target text replaced by the subsequent candidate, and other character strings may be displayed in different colors. Further, it is also possible to simply display the expression to be output without particularly distinguishing these character strings.

  When the output process (step S26 in FIG. 5) described with reference to FIGS. 9 to 11 ends, the process of the procedure in the example of FIG. 5 ends.

  The information processing apparatus 10 may accept an instruction from a user who has confirmed the display screen in the output process of step S26 in FIG. 5 and correct the processing target text in the electronic document by the expression on the display screen. For example, when the user selects an expression on the third line using the input device on the display screens in the examples of FIGS. 9 to 11, the input reception unit 120 of the information processing apparatus 10 converts the selected expression into a modified text generation unit. 180 is notified. Upon receiving this notification, the corrected text generation unit 180 generates the expression “a tumor is seen in the right lung” corresponding to the sentence on the third line, and generates the processing target text in the electronic document including the processing target text. Replace with the text you made. As a result, the original text to be processed in the electronic document “a high absorption area is seen in the right lung” is corrected to the character string “a tumor is seen in the right lung”.

  In the processing of the example of the embodiment described above, not only an expression in which a new word or phrase is inserted at the specified position of the processing target text, but also an expression in which a part of the character string after the specified position is replaced with another character string. Can also be generated. By outputting these expressions, candidates for correction of the text to be processed are presented to the user.

  The information processing apparatus 10 exemplified above is typically realized by executing a program describing functions or processing contents of each unit of the information processing apparatus 10 described above on a general-purpose computer. In the computer, for example, as shown in FIG. 12, a CPU (central processing unit) 80, a memory (primary storage) 82, various I / O (input / output) interfaces 84, and the like are connected via a bus 86 as hardware. Circuit configuration. In addition, a hard disk drive (HDD) 88, a disk drive 90 for reading various types of portable non-volatile recording media such as a CD, a DVD, and a flash memory is connected to the bus 86 via, for example, an I / O interface 84. Connected. Such a drive 88 or 90 functions as an external storage device for the memory. A program in which the processing content of the embodiment is described is stored in a fixed storage device such as the HDD 88 via a recording medium such as a CD or DVD or via a network, and is installed in a computer. The program stored in the fixed storage device is read into the memory and executed by the CPU, whereby the processing of the embodiment is realized.

  In the above, the embodiment of the example in which the information processing apparatus 10 is realized by one computer has been described. However, the various functions of the above-described example of the information processing apparatus 10 may be realized by being distributed to a plurality of computers. Good.

  10 Information processing device, 80 CPU, 82 Memory, 84 I / O interface, 86 bus, 88 HDD, 90 Disk drive, 100 Document storage unit, 110 Reference data storage unit, 120 Input reception unit, 130 Target text acquisition unit, 140 Subsequent character string estimation unit, 150 text decomposition unit, 160 linked expression evaluation unit, 170 output processing unit, 180 modified text generation unit, 200 display screen.

Claims (7)

A reception step for receiving a designated position that is a position in a target character string that is a character string to be processed;
Referencing a dictionary storage means storing a plurality of sentences, using a character string included up to the specified position in the target character string as a search key, and using the search key among the plurality of sentences stored in the dictionary storage means In a sentence including the character string, a guess step of inferring a character string following the character string of the search key as a candidate character string that is a candidate for a character string following the character string up to the specified position in the target character string;
An analysis step for performing morphological analysis on a character string after the specified position in the target character string;
A decomposing step of decomposing a character string after the specified position in the target character string into an element character string that is a character string corresponding to a component of a sentence based on the result of the morphological analysis ;
For each pair of the candidate character string and the element character string, an expression in which the character string in the target character string after the element character string is followed by the candidate character string is evaluated based on a predetermined evaluation rule. An evaluation step for obtaining an evaluation value representing the validity of the expression as a natural language;
Outputting the expression in which the evaluation value is equal to or greater than a preset threshold;
A program that causes a computer to execute. Generates a character string in which a character string after the specified position in the target character string is replaced with the expression selected by the input in response to a user input selecting one of the expressions output in the output step. Generating step,
The program according to claim 1, further causing the computer to execute the following. In the output step, when there are a plurality of expressions having the evaluation value equal to or greater than the threshold value, display control information for controlling the expression having a higher evaluation value to be displayed on the display unit with priority is output. To
The program according to claim 1 or 2, characterized in that In the evaluation step, the evaluation value is not obtained for the same expression as the expression included in the target character string.
The program according to any one of claims 1 to 3, wherein: In the output step, display control information for controlling the display means to display the candidate character string in the expression whose evaluation value is equal to or greater than the threshold and the character string in the target character string in a different manner is output. To
The program according to any one of claims 1 to 4, wherein: In the output step, the target character string is further arranged between the designated position of the target character string and the element character string included in the expression having the evaluation value equal to or greater than the threshold value. Outputting display control information for controlling the element character string to be displayed on the display means together with the expression,
The program according to any one of claims 1 to 5, wherein: Accepting means for accepting a designated position that is a position in a target character string that is a character string to be processed;
Referencing a dictionary storage means storing a plurality of sentences, using a character string included up to the specified position in the target character string as a search key, and using the search key among the plurality of sentences stored in the dictionary storage means In a sentence including the character string, a guess means for guessing a character string following the character string of the search key as a candidate character string that is a candidate for a character string following the character string up to the specified position in the target character string,
Analyzing means for performing morphological analysis on a character string after the specified position in the target character string;
Decomposing means for decomposing a character string after the specified position in the target character string into an element character string that is a character string corresponding to a sentence component, based on the result of the morphological analysis ;
For each set of the candidate character string and the element character string, in accordance with a predetermined evaluation rule, evaluate the expression that the candidate character string is followed by the character string in the target character string after the element character string, An evaluation means for obtaining an evaluation value representing the validity of the expression as a natural language;
Output means for outputting the expression in which the evaluation value is equal to or greater than a preset threshold;
An information processing apparatus comprising:

Images