JP3219527B2 - Information processing device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an information processing apparatus having a function of understanding the meaning of input information.

[0002]

2. Description of the Related Art Conventionally, apparatuses that execute processing based on information input in a natural language have been put into practical use in fields such as database search. Further, recently, handwriting input with a pen has been attracting attention as an input means instead of a keyboard, and devices that understand the meaning of a character string input by handwriting and execute processing have been developed (for example, “Nikkei Byte” 1992). July issue, P96-P97).

In this type of apparatus, the character string obtained as a result of the recognition processing is further subjected to syntax analysis and semantic analysis to understand the meaning of the input and execute the processing. For example, if you write "Output document A to printer" on the input device,
The contents of a document named document A are output from a predetermined printer. By the way, when requesting a process to another person by writing a memo or the like, a simple expression including symbols such as arrows as shown in FIG. 2A is often used.

However, such a conventional language processing apparatus that targets only characters cannot understand the meaning of information in which characters and symbols are mixed. On the other hand, as a device that understands the meaning from information in which characters and symbols are mixed and executes processing, it calculates a handwritten mathematical formula input in a handwriting format as described in Japanese Patent Application Laid-Open No. 4-76756 (G06F15 / 20). An apparatus has been proposed.

[0005] In this type of device, the meaning of the symbols used is limited, and thus, the input numeral strings and symbols are merely replaced with appropriate forms based on individual positions. However, in an instruction composed of a natural language and a symbol as shown in FIG. 2A, the meaning of the symbol used is not necessarily limited as in the case of a mathematical expression. For example, the arrow symbol in FIG. 2A means "output (material A) to (printer)".

However, if the same arrow symbol is input as shown in FIG. 3 (a), it means "Please send ~ (phone) to ~ (Suzuki)".

[0007]

As described above, one symbol may be used in a variety of meanings depending on the situation in the same manner as a general word, but in the related art, such a broad meaning is used. Cannot correctly understand the instructions that include the symbols used in The present invention has been made in consideration of the above-described circumstances, and realizes an information processing apparatus that correctly understands the meaning of information including characters and symbols including symbols used in various meanings.

[0008]

An information processing apparatus according to the present invention comprises an input means (1) for inputting input information consisting of characters and symbols, and an input information input by the input means (1). Word string generating means (6) for converting to a word string equivalent to
Word dictionary section containing the meaning and grammatical attributes of words (8)
And a semantic interpreting means (9) for interpreting the meaning of the input information by analyzing the word string generated by the word string generating means (6) with reference to the word dictionary section (8). .

In the information processing apparatus according to the present invention, in the word string generating means (6), when the input information includes a symbol, the input information has a meaning and a grammatical attribute corresponding to the symbol. A means for converting to a word string including virtual words and a conversion dictionary unit (7) for conversion, wherein the word dictionary unit (8) converts the meaning and grammatical attributes of the virtual words to general words. It is characterized by having been recorded with.

Further, in the document processing apparatus of the present invention, the input means (1) includes a coordinate input means (1) for inputting handwriting coordinates, and a handwriting coordinate input by the coordinate input means (1). Recognition means for recognizing characters and symbols from data (4)
And a writing position detecting means (5) for detecting a writing position of the character and the symbol.

[0011]

In the information processing apparatus of the present invention, the word string generating means (6) converts input information consisting of characters and symbols into a word string containing virtual words corresponding to symbols, and interprets the meaning (9). By analyzing word strings by referring to the word dictionary section (8), which stores the meaning and grammatical attributes of the virtual words in advance, characters and symbols are treated in the same language system and the meaning of input information To understand the.

[0012]

FIG. 1 is a block diagram showing a functional configuration of an information processing apparatus according to an embodiment of the present invention. In the figure, (1)
Is a coordinate input device. This coordinate input device (1) is mainly composed of a tablet, and the coordinates (x,
y), and detects time-series data indicating pen handwriting (hereinafter, referred to as
Handwriting coordinate data).

(2) is a control unit. This control unit (2)
It is mainly composed of a CPU and receives handwriting coordinate data and performs various controls such as inputting and outputting data between the devices shown in FIG. (3) is a character / symbol discriminating device. This character / symbol discriminating device (3) is provided with a coordinate input device via a control unit (2).
Based on the handwriting coordinate data input from (1), it is determined whether the stroke set constituting the writing information is a character or a symbol, and the result is written together with the handwriting coordinate data into a character / symbol recognition device (4). And it outputs to the writing position detecting device (5).

(4) is a character / symbol recognition device. This character / symbol recognition device (4) comprises a known character recognition processor and a known symbol recognition processor. The character / symbol recognition device (4) performs character recognition processing on a stroke set determined as a character in the character / symbol determination device (3), and performs a symbol recognition process on a stroke set determined as a symbol. Perform recognition processing. And this character / symbol recognition device (4)
Returns a character code or a symbol code as a recognition result to the control unit (2) via the character / symbol discrimination device (3).

(5) is a writing position detecting device. This writing position detection device (5) obtains the writing position coordinates (x, y) of each stroke set from the handwriting coordinate data, and sends the coordinate values to the control unit (2) via the character / symbol discrimination device (3). return. (6)
It is a word string generation device. The word string generation device (6) is configured to recognize a recognition result input from a character / symbol recognition device (4) and a writing position detection device (5) via a control unit (2) and a character / symbol discrimination device (3). Based on the writing position coordinates, a word string corresponding to the writing information is generated with reference to the word dictionary section (8).

At the time of this generation, the word string generation device (6)
If the input includes a symbol, first arrange the respective elements of the character string and the symbol in an appropriate order according to the type of the symbol, and then, by referring to the conversion dictionary unit (7), Convert to a character string containing virtual words corresponding to symbols. Finally, the character string is divided into words using the word dictionary unit (8) to generate a word string.

(7) is a conversion dictionary unit. The conversion dictionary unit (7) is mainly composed of a ROM, and is information for converting a character string including a symbol into a character string including a virtual word corresponding to the symbol registered in the word dictionary unit (8). Is recorded. (8) is a word dictionary section. This word dictionary (8)
A virtual word mainly composed of a ROM and having a general word and a meaning and a grammatical attribute equivalent to a symbol is registered.

(9) is a language analyzer. The language analysis device (9) performs a syntax analysis and a semantic analysis on the word string generated by the word string generation device (6), obtains a semantic structure of the handwritten information, and outputs it to the execution unit (10). (10) is an execution unit. The execution unit (10) executes necessary processing based on the meaning of the handwritten information obtained by the language analysis device (9).

FIG. 2 is a diagram for explaining the operation of the information processing apparatus according to one embodiment of the present invention. FIG. 2A shows a state in which writing information has been input in the coordinate input device (1). (21), (22), and (23) are input writing stroke sets, respectively. (24) is a character entry frame previously shown on the writing surface. The handwriting coordinate data corresponding to the writing information is first sent to the character / symbol discriminating device (3) via the control unit (2).

The character / symbol determining device (3) determines a character / symbol from the input handwriting coordinate data. The distinction between a character and a symbol is made based on the size of the stroke set for the character entry frame. That is, it is determined that a character written in one entry frame is a character, and a character written over two or more entry frames is a symbol. In this example, the stroke sets (21) and (23) are determined as character strings, and the stroke set (22) is determined as a symbol.

The character / symbol recognition device (4) performs a recognition process according to the judgment result of the character / symbol discrimination device (3). The stroke set (21) is composed of the character string “Document A” and the stroke set (22).
Is recognized as a “right-pointing arrow symbol”, and the stroke set (23) is recognized as a character string “printer”. The recognition result is output to the word string generator (6) via the character / symbol discriminator (3) and the controller (2).

The writing position detecting device (5) includes (21), (22), and (23)
The coordinates of the upper left point of each stroke set are obtained, and the coordinate values are sent to the word string generation device (6) via the character / symbol determination device (3) and the control unit (2). The word string generation device (6)
A word string is generated from the recognition result obtained by the character / symbol recognition device (4) and the writing position coordinates obtained by the writing position detection device (5). In this example, since the stroke set (22) is a “right-pointing arrow symbol”, first the sequence {(character string to the left of the arrow symbol), (arrow symbol), (character string to the right of the arrow symbol)} A string is generated. FIG. 2B shows the character string.

Next, with reference to the conversion dictionary section (7), replacement and supplementation of a symbol portion with a virtual word having a meaning and a grammatical attribute corresponding to the symbol are performed.
FIG. 4 shows a part of the contents of the conversion dictionary unit (7). In FIG. 4, (S1) and (S2) are arbitrary character strings, <P1>, <P2
> And <V> are virtual words registered in the word dictionary section (8). The dictionary content of these words is shown in FIG. 5, along with other words used in FIGS. The converted character string is further divided into words by referring to the word dictionary section (8), and is output as a word string to the language analyzer (9).

Thus, a character string including an arrow symbol as shown in FIG. 2B is converted as shown in FIG. 2C. Language analyzer
In (9), first, a syntax analysis is performed on the word string input from the word string generation device (6) based on the grammar rules, and a phrase structure tree as shown in FIG. 2 (d) is created. Next, the semantic relationship between the components is analyzed with respect to the obtained phrase structure tree.

The processing of the semantic analysis is performed while referring to the grammatical attributes of each word in the word dictionary (8). Verb <V>
By referring to the concept / particle-connection relation attribute, the case element “document A <P1>” becomes the target case of verb <V>,
"<P2>" is the target case of verb <V>, and the meaning of verb <V> can be determined to be "output". As a result of the semantic analysis, a semantic structure as shown in FIG. The obtained semantic structure is sent to the execution unit (10), and the contents of the document A are output from a predetermined printer.

Next, the operation for the input shown in FIG. 3A will be described below. In the figure, (31), (32), and (33) are writing stroke sets input from the coordinate input device (1). As in the previous case, the stroke set (31) is a character string "TEL" and the stroke set (32) is a "right-pointing arrow sign" by the character / symbol discriminating device (3) and the character / symbol recognizing device (4). , The stroke set (33) is recognized as a character string “Mr. Suzuki”. In addition, by the writing position detection device (5),
The writing position coordinates of each stroke set are obtained.

The word string generator (6) uses these results to create the character string shown in FIG. 3 (b), and then generates the word string shown in FIG. ). The language analyzer (9) obtains the phrase structure tree of FIG.
Further, as a result of the semantic analysis, the semantic structure of FIG. 3 (e) is obtained. In this example, the meaning of the verb <V> is “transmit” from the semantic relationship between the components.

As described above, by using the present invention, the meaning of information including symbols used in various meanings depending on the scene, such as the arrow symbols used in FIGS. 2 (a) and 3 (a), can be obtained. Can be correctly understood. In the present embodiment, a method using a character entry frame for character / symbol discrimination in the character / symbol discrimination device (3) is adopted, but this does not limit the features of the present invention, and May be used as appropriate.

The scope of application of the present invention is not limited to only handwriting input. For example, instead of each of the devices (1) to (5) of the present embodiment, a code string is input and transmitted by a keyboard, a touch panel, or the like. The device can be used as the input means. Further, the execution unit (10) can be configured by using various devices that execute processing based on input, and for example, besides outputting and communicating documents described in the present embodiment, for example, searching for an address book A wide range of applications are possible, such as schedule management.

[0030]

According to the present invention, when analyzing the meaning of information consisting of characters and symbols, both are handled in the same linguistic system, and thus include symbols used in various meanings depending on the scene. It is possible to correctly understand the meaning of the information and to flexibly deal with various inputs.

In particular, in an information processing apparatus based on handwriting input, the effect of the present invention is extremely high because it is possible to input in a simple and natural expression unique to handwriting.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a functional configuration of an information processing apparatus according to an embodiment of the present invention.

FIG. 2 is a diagram illustrating an operation example of an information processing apparatus according to an embodiment of the present invention.

FIG. 3 is a diagram illustrating an operation example of the information processing apparatus according to the embodiment of the present invention.

FIG. 4 is a diagram showing the contents of a conversion dictionary unit of the information processing apparatus according to one embodiment of the present invention.

FIG. 5 is a diagram showing the contents of a word dictionary section of the information processing apparatus according to one embodiment of the present invention.

[Explanation of symbols]

 (1) Coordinate input device (input means), (2) control unit, (3) character / symbol discrimination device, (4) character / symbol recognition device (recognition means), (5) writing position detection device (writing position detection (6) Word string generator (word string generator), (7) conversion dictionary, (8) word dictionary, (9) language analyzer (meaning interpreter), (10) execution unit (process) Execution part).

 ──────────────────────────────────────────────────続 き Continued on the front page (72) Inventor Hideto Yamamoto 2-18-18 Keihanhondori, Moriguchi-shi, Osaka Sanyo Electric Co., Ltd. (72) Inventor Hiroshi Horii 2--18 Keihanhondori, Moriguchi-shi, Osaka (56) References JP-A-4-355871 (JP, A) JP-A-4-119466 (JP, A)

Claims (2)

(57) [Claims] An input means (1) for inputting an input character string consisting of a character and a specific symbol; information before and after a specific symbol on the input character string and information on the specific symbol before conversion; A conversion dictionary unit (7) in which the preceding and following information and the virtual word and the information related to the arrangement are registered in association with each other; and referring to the conversion dictionary unit (7), the input character string input by the input means (1) A word string generating means (6) for converting into a character string including a virtual word; a word dictionary section (8) in which the meaning, concept, and grammatical attributes of the word and the virtual word are registered; ), The word string generating means (6)
A language analysis unit that analyzes the word string generated in step (a) and selects one from a plurality of meanings of the virtual word in which the specific symbol has been converted, based on the meanings and concepts of words before and after the specific symbol ( 9) An information processing apparatus comprising: 2. The input means (1) comprises: coordinate input means (1) for inputting handwriting coordinates; and recognition for recognizing characters and symbols from handwriting coordinate data input by the coordinate input means (1). Means (4), and a writing position detecting means (5) for detecting a writing position of the character and the symbol. The word string generating means (6) converts the input character string input by the input means (1). 2. The information processing apparatus according to claim 1, further comprising means for converting a word string equivalent to the input character string.