KR101427040B1 - transaction information input system for a smart device using speech recognition and information input method for asset management using thereof - Google Patents

Abstract

The present invention relates to a transaction data system for a smart device using voice recognition and a data input method for asset management using the same. More specifically, the system includes a voice input unit which receives the voice of a user when the user vocalizes a sound of one among the date, time, category, payment method, amount, account, type, or details of a transaction in order to input housekeeping book data into a smart device; a voice recognition engine which recognizes the voice inputted into the voice input unit and generates voice data from the recognized content; a voice analysis unit which analyzes the meaning of the transaction based on the voice data generated by the voice recognition engine, extracts the transaction data of the transaction, and stores the extracted transaction data into a linked storage unit; and a transaction data display unit which displays the transaction data extracted by the voice analysis unit onto the display means of the smart device.

Description

Technical Field [0001] The present invention relates to a transaction information input system for a smart device using voice recognition and an information input method for asset management using the same,

The present invention relates to a transaction information input system for a smart device using voice recognition and an information input method for asset management using the same, and more particularly, A voice input unit for inputting a voice of a user who uttered at least one of a time, a category, a payment means, an amount, a customer, a type, and a detailed content; A voice recognition engine for recognizing a voice input to the voice input unit and generating voice information composed of the recognized voice; A voice analyzer for analyzing the meaning of the transaction based on the voice information generated by the voice recognition engine to extract transaction information of the transaction and storing the extracted transaction information in a correlated storage unit; And a transaction information display unit for displaying the transaction information extracted by the voice analysis unit on the display unit of the smart device.

A smart household account service method and system of Korean Patent Application Laid-Open No. 10-2012-138175 is known as a household account application that runs on a smart device such as a smart phone.

The above prior art analyzes the authorization history text messages (SMS, LMS, MMS) sent from the card issuer server on the network of the communication company when making a payment using a credit card or a credit card, And it is possible to make it possible to inquire the household part through a web (Web) which is connected with a dedicated application installed on a user's smart phone or a general PC, It is a system that enables the user terminal to inquire the household part through sync.

However, in the above-mentioned Smart Household Service method, there is a simplicity in that the household banking department automatically creates the transaction using the financial card. However, when entering a transaction that does not use the financial card, such as a cash transaction, It is inconvenient to input the data.

Also, a method of automatically recording a household account using a mobile social network service of Korean Patent Laid-Open Publication No. 10-2012-20679 is also known. This prior art can automatically transmit and store expenses to a PIM server in a virtual space using mobile SNS Is a household system application system.

However, this prior art also has a problem in that it is troublesome to transmit expenditure contents to the PIM server all the time to the SNS, so that the convenience of the user may be degraded as in the above-described prior art.

The present invention has been made in order to solve the problems of the related art as described above, and it is an object of the present invention to provide a smart device capable of voice recognition, in which a user can input transaction information necessary for household account creation, income / expenditure management, The present invention provides a transaction information input system for a smart device using voice recognition that can increase user's convenience in inputting transaction information in preparation of a household account, and an information input method for asset management using the same.

According to an aspect of the present invention, there is provided a transaction information input system for a smart device using voice recognition, comprising: a transaction information inputting system (1) of a smart device (1) having a display means (2) A voice input unit 10 for receiving a user voice uttered for inputting transaction information from the microphone 3 of the smart device 1; A voice recognition engine 20 for recognizing the voice input to the voice input unit 10 and generating voice information composed of the recognized voice; A voice analysis unit 30 for extracting transaction information of the transaction based on the voice information generated by the voice recognition engine 20 and storing the extracted transaction information in the interlocked storage unit 50; A transaction information display unit (40) for displaying the transaction information extracted by the voice analysis unit (30) on the display means (2) of the smart device (1); And a control unit.

      Preferably, the information for inputting the transaction information input to the voice input unit 10 is at least one of a category, an asset, a customer, and a detailed content.

Here, the speech analyzing unit 30 may search whether the words included in the voice information match a category stored in the storage unit 50 or an item of an asset or an item of a customer, If the words that match the stored category items are included in the voice information, the matching category items are assigned as categories, and if the voice information includes words that match the stored asset items, If the asset item is designated as an asset and a word corresponding to the stored account item is included in the voice information, it is preferable to designate the matching account item as a business partner.

Here, if the transferred voice information does not include a word that matches a category or an asset previously stored in the voice information, or a word corresponding to the item of the customer, the voice analysis unit 30 stores voice information It searches the category or the assets included in the information or at least one of the details or the details of the transactions and matches the category or the information of the specified category or the asset or the customer or the details in the past transaction information with a large number of matching items, It is desirable to designate it as a supplier or detail.

Preferably, the information for inputting the transaction information input to the voice input unit 10 is at least one of a date, a time, and an amount of money.

Here, the voice analyzer 30 converts the date information into a date when the voice information includes numeric or numeric characters and includes a date unit such as year, month, and day, extracts date information of the corresponding transaction, It is preferable that the date on which the voice information is input is extracted as the date information.

Here, it is preferable that the voice analyzing unit 30 extracts the voice information as a money item when the number is confirmed and the word of the unit of money is included.

In addition, the information for inputting transaction information, which is input to the voice input unit 10, is preferably information on a transaction type or detailed contents.

Here, when the voice information includes the word import, expenditure or transfer, the voice analysis unit 30 extracts a transaction type as a corresponding word, and when the word import, expenditure or transfer is not included, It is desirable to search whether an included word exists in an item of a category stored in the storage unit and, if so, extract the transaction type of the category as a transaction type of the inputted voice information.

In another aspect, the present invention provides an information input method for asset management using a transaction information input system for a smart device having a speech recognition engine for generating voice information and a storage unit for storing information for asset management, A user voice input step (S100) of inputting a voice of the user into the smart device when the information of the transaction is generated; (Step S 200) of generating voice information for recognizing a voice input by using the voice recognition engine of the smart device and generating voice information composed of the recognized contents; An extraction step (S 300) of extracting transaction information of the transaction based on the voice information generated by the speech recognition engine and storing the extracted transaction information in the interlocked storage unit; And displaying the extracted transaction information on the display means of the smart device (Step S 400).

Preferably, the step of extracting transaction information (S 300) includes at least one of a category extracting step (S 310), an asset extracting step (S 320), and a customer extracting step (S 330).

Here, it is searched whether there is a word that matches any one item of the category or the asset stored in the storage unit of the voice information, or the item of the account, and if a word matching the category item stored in the search result is voice information , The category item is designated as a category, and if a word corresponding to the stored asset item is included in the voice information, the asset item is designated as an asset, If the matching word is included in the voice information, it is preferable to designate the corresponding account item as a customer.

Here, if the transferred voice information does not include a word stored in the category or the asset or the account of the customer, the at least one word among the words included in the voice information among the past transaction information stored in the storage unit And to designate the information of the specified category or asset or business partner or detail on the past transaction information having a large number of matching items and high accuracy as the category or the asset or the business partner or the details.

If none of the words included in the transmitted voice information matches any of the past transaction information, it is determined whether or not a category or an asset most appropriate to the word from the past transaction information stored in the server connected to the smart device Or recommending the information of the customer to the user

Herein, when extracting the customer, it is preferable that the location information of the customer is extracted together and the customer close to the location information of the user at the time when the user inputs the voice is automatically extracted.

The transaction information extracting step S 300 may include at least one of a date and time extracting step S 340, an amount extracting step S 350 and a transaction type extracting step S 360 .

The date and time extracting step (S 340) includes a step of extracting the date information of the corresponding transaction by converting the date and time into a date if the voice information includes a numeric or numeric character and a date unit such as year, month, And if the year, month, or day of the date information is omitted, the date on which the voice information is input is extracted as the date information.

In addition, if the number of the voice information is confirmed and the word of the unit of money is included in the voice information, the amount extracting step (S 350) extracts the number as a monetary amount, ) Is preferably combined with a number and extracted as a monetary item.

Here, if the amount of money can not be extracted from the transmitted voice information, the voice analysis unit 30 determines that at least one word among the words included in the voice information among the past transaction information stored in the storage unit 50 exists And if there is a word matching the past transaction information, it is preferable to extract the amount of money from the past past transaction information.

In addition, the transaction type extracting step (S 360) extracts a transaction type of a corresponding word when the voice information includes the word import, expenditure or transfer, and when the word import, expenditure or transfer is not included , It is desirable to search whether the words included in the voice information exist in the category of the category stored in the storage unit and, if so, to extract the transaction type of the category as the transaction type of the inputted voice information.

If there is no category item, the voice analysis unit 30 searches the past transaction information stored in the storage unit 50 to see if there is at least one word among the words included in the voice information, When there is a word matching the transaction information, it is preferable to extract the transaction type from the past transaction information.

In another aspect, the present invention relates to a recording medium containing a program to be executed, including a method for inputting information for managing the asset.

The transaction information input system for a smart device using the speech recognition of the present invention having the above configuration and the information input method for asset management using the same provide transaction information necessary for creating a household account, It is effective to increase the convenience of creating a household account, managing income / expenditure, and asset management.

In addition, according to the embodiment, when a user inputs a date, a time, a category, an account (cash, bank, card, etc.), an amount, a customer, a transaction type, The transaction information of each input item is generated based on the past transaction information stored in the past, and the transaction information is additionally confirmed to the user, thereby improving the accuracy and convenience of the transaction information .

1 is a block diagram of a transaction information input system according to the present invention,
FIGS. 2A to 2E illustrate an embodiment of an input window displayed on a display means of a smart device to which a transaction information system of the present invention is applied,
3 is a flowchart of an information input method for asset management according to the present invention,
FIGS. 4A to 4F are detailed flowcharts of the step of extracting transaction information in FIG. 3,
5A and 5D are views showing an embodiment of an input window displayed on the display means of the smart device showing the information input method of the present invention.

Hereinafter, a configuration of a transaction information input system for a smart device using voice recognition of the present invention and an information input method for asset management using the same will be described in detail with reference to the accompanying drawings.

It is to be noted, however, that the disclosed drawings are provided as examples for allowing a person skilled in the art to sufficiently convey the spirit of the present invention. Accordingly, the present invention is not limited to the following drawings, but may be embodied in other forms.

In addition, unless otherwise defined, the terms used in the description of the present invention have the same meaning as commonly understood by one of ordinary skill in the art to which the present invention belongs. In the following description and the accompanying drawings, A detailed description of known functions and configurations that may be unnecessarily blurred is omitted.

1 is a block diagram illustrating a configuration of a transaction information input system for a smart device using speech recognition of the present invention.

In order to solve the inconvenience of manually inputting transaction information in a household system or the like in the past, the system of the household system of the present invention is capable of inputting corresponding transaction information by voice. Particularly, ). ≪ / RTI >

The smart device 1 may be any known mobile communication device such as a mobile phone, a PDA, or a tablet PC. However, it is preferable that an application program (application program) It is desirable to be a so-called 'smart phone' or 'tablet PC' which can be downloaded and executed.

The smart device 1 is also provided with a screen display means 2, a microphone 3 for voice input, and a key input module 4 for inputting letters or numbers.

The display means 2 may be a conventional monitor equipped with a liquid crystal display, an LED matrix display, a cathode-ray tube or an LCD panel, but is preferably a pressure-sensitive or capacitive- A display panel is employed.

In order to allow the user to input the transaction information information into the smart device 1, the system of the present invention as described above may include information for inputting the transaction information, such as date, time, category, asset, amount, A voice input unit (10) for receiving a voice of a user uttered from the microphone (3) of the smart device (1) when at least one item is uttered; A voice recognition engine 20 for recognizing the voice input to the voice input unit 10 and generating voice information composed of the recognized voice; A voice analysis unit 30 for extracting transaction information of the transaction based on the voice information generated by the voice recognition engine 20 and storing the extracted transaction information in the interlocked storage unit 50; A transaction information display unit (40) for displaying the transaction information of the transaction extracted by the voice analysis unit (30) on the display means (2) of the smart device (1); .

The smart device 1 communicates with an external server 100 through a communication network such as a wireless Internet to transmit and receive information. To this end, the smart device 1 may employ a wireless communication module capable of connecting to a communication network. The wireless communication module may be a Zigbee, a Bluetooth, an Infrared Data Association (IRDA) (Home Phoneline Networking Alliance), Shared Wireless Access Protocol (SWAP), Wireless Internet Network (WLAN), and the like.

In the embodiment of the present invention, the 'category' item among the items of the corresponding transaction received from the voice input unit 10 means the purpose of the corresponding transaction. For example, the categories may include meals, athletics, beauty, transportation, hospitals, salaries, insurance, and the like.

In this case, in order to facilitate the use of the transaction information of the present invention, it is preferable to previously set the category of the category in the system of the present invention. For example, in general, such as a meal, a movement, a beauty, a traffic, a hospital, And stores the items that can be applied to a plurality of users as a category in the storage unit 50 in advance.

The type of the category can be directly input by the user, and the type of the category can be inputted through the key input module 4 of the smart device 1 and stored in the storage unit 50 .

In addition, the category may be set as an upper category indicating a comprehensive purpose of the transaction and a lower category subdivided as the content of the upper category. The display of the smart device 1 of the system of the present invention shown in Figs. 2A and 2B When the upper category is' transportation fee 'as in the embodiment of the window displayed in the means 2, the subcategories of the' transportation fee 'category include' bus', 'subway', 'taxi', ' Item can be registered.

The 'date' and 'time' of the corresponding transaction automatically designates the date and time when the voice is inputted through the voice input unit 10.

The "assets" of the transaction are cash (cash stored in a purse or safe) Refers to an item for which a user's own goods such as a bank account, a credit card, a check card, a transportation card, a loan account, an insurance account, and a real estate transaction such as a demand deposit account, a term deposit account, and a regular account are traded.

The item of 'asset' is an item of the asset as shown in the input window displayed on the display means 2 of the smart device 1 to which the system of the present invention is applied, as shown in FIG. 2C, , Cash, bank, card, check card, savings, investment, minus bankbook, loan, insurance and so on.

At this time, the 'asset' item may be set as an upper asset item which comprehensively indicates the contents of the asset and a lower asset item which subdivides the contents of the higher asset.

The 'amount' item of the above transaction indicates the amount of the transaction amount.

The 'customer' item of the transaction indicates the information of the counter party. FIG. 2D shows an embodiment of the 'customer' input window displayed on the display means 2 of the smart device 1 to which the system of the present invention is applied.

At this time, the 'customer' item may be set as an upper customer who comprehensively represents the content of the asset and a lower customer who subdivides the upper customer.

The 'type' item of the corresponding transaction includes an 'income' item in which the transaction is transferred to the user's asset item, an item 'expense' (withdrawal) in which the amount is withdrawn from the asset item, and an item It points to the 'transfer' items to be deposited.

FIG. 2E shows an example of an input window of a 'type' item displayed on the display means 2 of the smart device 1 to which the system of the present invention is applied. In the illustrated embodiment, The embodiments of the remaining 'import' and 'transfer' items may also be configured similar to the embodiment of the 'spend' item.

The 'details' of the corresponding transaction may include other transactions that do not correspond to any of the date, time, category, asset, amount, customer, or type of the transaction It refers to the item and contains specific information about the transaction.

The speech recognition engine 20 is a part in which a known speech recognition program configured on the basis of a speech recognition methodology is driven. The speech recognition engine 20 recognizes a speech inputted by the user and generates information to generate speech information of the inputted speech. Since the speech recognition by the speech recognition program can be performed through a well-known process, a detailed description thereof will be omitted. A speech recognition engine provided in the smart device may be used, or may be separately manufactured, As shown in FIG.

The voice analysis unit 30 analyzes the meaning of the transaction based on the voice information generated by the voice recognition engine 20 and extracts transaction information of the transaction. The transaction information is extracted from the voice information by operation.

1) Extraction of category items of the transaction

The voice analysis unit 30 determines whether or not a word included in the voice information matches any one category item of the category stored in the storage unit 50 based on the voice information transmitted from the voice recognition engine 20 And if a word coinciding with the category item stored in the search result is included in the voice information, the category item is designated as the category of the received voice information. In this case, when a word included in the voice information is searched, the searched word can be searched with or without spacing between the words, and the searched searched word can be searched for within the error range.

If the words included in the voice information coincide with the items corresponding to the items of the lower category, the voice analysis unit 30 is automatically designated as the category corresponding to the upper category of the lower category.

On the other hand, if the transmitted voice information does not include words that match the categories stored in advance, the voice analysis unit 30 can infer the category from the words included in the voice information.

That is, it is checked whether at least one word among the words included in the voice information exists in the past transaction information stored in the storage unit 50. If there is a word matching the past transaction information, Designates a category designated as a category item.

For example, when voice information called 'cash Americano 5,000 won Starbucks' is delivered, 'cash' (asset), 'americano' (details), '5,000 won' (amount), 'Starbucks' The voice analysis unit 30 of the embodiment of the present invention does not include the word clearly indicating the category in the voice information as described above. Therefore, the voice analysis unit 30 determines the category of the past transaction information , It is searched whether transaction information having the same item as the item of 'Americano' (details) or 'Starbucks' (customer) or 'Cash' (asset) exists. If there is past transaction information in which specific information matches, the category of the past transaction information is extracted.

However, if there is no word in any of the words included in the transmitted voice information that matches the past transaction information, the voice analysis unit 30 does not specify the category of the voice information, A category may be specified, or a category set by the user may be specified for the case where there is no category, or the category most appropriate to the word may be recommended to the user.

The voice analysis unit 30 transmits voice information to the external server 100 using the wireless communication module built in the smart device 1, The transaction information of the user's own transaction information stored in the server or the transaction information of another user is checked for the existence of at least one word among the words included in the voice information. If there is a word corresponding to the past transaction information, It is possible to recommend the designated category as a category item to the user on the past transaction information.

2) Extraction of asset items of corresponding transaction

     As described above, the 'asset' of the embodiment of the present invention is a property of a user's own account such as a demand deposit account, a term deposit account, a bank account such as a regular account, a credit card, a check card, a transportation card, a loan account, Indicates the item for which the goods are traded.

Therefore, based on the voice information transmitted from the voice recognition engine 20, the voice analysis unit 30 can extract any one of the asset items previously stored in the storage unit 50 among the words included in the voice information, If a word matching the previously stored asset item is included in the voice information, the corresponding asset item is designated as the asset of the voice information received. In this case, when a word included in the voice information is searched, the searched word can be searched with or without spacing between the words, and the searched searched word can be searched for within the error range.

If the words included in the voice information coincide with the items corresponding to the items of the lower asset, the voice analysis unit 30 can automatically extract the transaction information as assets corresponding to the upper assets of the lower asset.

On the other hand, if the transferred voice information does not include a word matching the stored items of the asset, the voice analysis unit 30 can deduce an asset item from the words included in the voice information.

That is, it is checked whether at least one word among the words included in the voice information exists in the past transaction information stored in the storage unit 50. If there is a word matching the past transaction information, And assigns the specified asset as an asset item.

However, if there is no word in any of the words included in the delivered voice information that matches the past transaction information, the voice analysis unit 30 does not specify the asset of the voice information, You specify the asset, specify the asset that the user has set for the asset in the absence, or recommend the asset item that best fits the word.

3) Extracting the item of the customer of the transaction

Based on the voice information transmitted from the voice recognition engine 20, the voice analyzing unit 30 extracts, from among the words included in the voice information, any one of the account items stored in the storage unit 50 And if a word corresponding to the stored customer item is included in the voice information, the customer item is designated as the customer of the received voice information.

In this case, when a word included in the voice information is searched, the searched word can be searched with or without spacing between the words, and the searched searched word can be searched for within the error range.

If the words included in the voice information coincide with the items corresponding to the items of the lower customer, the voice analyzer 30 can automatically extract the transaction information corresponding to the upper customer of the lower customer.

On the other hand, if the transferred voice information does not include a word matching the pre-stored items of the customer, the voice analysis unit 30 can infer the customer item from the words included in the voice information.

That is, it is checked whether at least one word among the words included in the voice information exists in the past transaction information stored in the storage unit 50. If there is a word matching the past transaction information, Is designated as a customer item.

However, if there is no word in any of the words included in the transmitted voice information that matches the past transaction information, the voice analysis unit 30 does not designate the customer of the voice information, If a customer is specified, or if there is no customer, the customer is designated as the customer, or the customer is referred to the most appropriate customer item.

As a recommendation method, voice information is transmitted to an external server 100 as described above, and a customer item is extracted from past transaction information of the user or transaction information of another user stored in the server 100, .

When the user inputs a voice, the location information of the customer is extracted from the location information receiving module such as a GPS terminal (not shown) built in the smart device 1, And extracts a customer corresponding to the position information.

4) Extract date and time items of the transaction

The voice analysis unit 30 is configured to calculate the voice information based on the voice information transmitted from the voice recognition engine 20 in the form of a number such as a number or a zero, a y, a, a, a, a, a, (Hereinafter, referred to as 'numeric character'), and when a date unit such as year, month, and day is included, it is converted into a date and date information of the corresponding transaction is extracted.

At this time, if any one of year, month, and day of the date information is omitted, date information is automatically extracted based on the date when the voice information is input. For example, when only the date is confirmed from the voice information, the year and the month at the input time of the voice information are automatically extracted. If only the month and date are confirmed from the voice information, If the year, month, and day are not confirmed from the voice information, the voice information is automatically extracted as the date of input of the voice information.

In addition, when the voice information includes words (yesterday, yesterday, the day before, the day after tomorrow, the day after tomorrow, the day after tomorrow, etc.) indicating the date, the word corresponding to the date is converted into the corresponding date and extracted. That is, if the date is not extracted from the audio information in the same manner as described above, it is determined whether or not other words indicating the date, that is, yesterday, today, tomorrow, yesterday, yesterday, , And compares the extracted date information with the date when the voice information is input, thereby extracting the date information.

If the voice information includes numeric or numeric characters and a word in time units such as hour, minute, and second is included in the voice information based on the voice information transmitted from the voice recognition engine 20, And extracts it as the time when the corresponding transaction information is made.

At this time, if any one of hour, minute, and second is omitted, time information is automatically extracted based on the date on which the voice information is input. For example, when only 'hour' is confirmed from the voice information, 'minutes' and 'seconds' at the input time of the voice information are automatically extracted. If only 'minute' is confirmed, Is automatically extracted at the input time point of the voice information when the hour, minute, and second are not confirmed from the voice information.

In addition, when the voice information includes a word indicating the time before or after the time such as 'before' or 'after' (for example, 30 minutes, 30 minutes, And extracted.

5) Extraction of the amount item of the transaction

If the number is confirmed in the voice information and the word of the currency (circle, dollar, yen, etc.) is included in the voice information based on the voice information transmitted from the voice recognition engine 20, the voice analysis unit 30 extracts do.

In addition, the voice information includes a unit character (e.g., billion, million, thousand, hundred, ten, etc.) indicating the number of digits or digits and the number of digits of the currency but does not include the unit of currency , The voice analysis unit 30 can deduce the unit of money of money from the words included in the voice information.

In the unit of currency used for the search, the stored money for use is set and used in the storage unit 50. According to an embodiment of the present invention, among the stored past transaction information, among the words included in the voice information, It is possible to search for the existence of a word indicating a currency unit and to extract a money item in units of the corresponding currency when there is a word indicating the currency unit of the amount in the past transaction information.

However, when only a number that does not include a denomination is extracted, it is preferable that the denomination of the denomination of the denomination of the denomination of the denomination of the denomination of the denomination of the denomination of the denomination of the denomination of the denomination. According to an embodiment, it is possible to select one of the denominations of money in the past transaction information searched and extract it as a money item by combining it with a number.

If the amount of money can not be extracted from the transferred voice information, the voice analysis unit 30 determines whether at least one word among the words included in the voice information among the past transaction information stored in the storage unit 50 exists If there is a word matching the past transaction information, it is preferable to extract the amount of money as the amount of money on the past past transaction information.

On the other hand, there are cases where common words include numeric characters. For example, in the word 'congratulatory letter', the numeral character 'ga' is included, and in the word 'ginseng', the numeral character 'ja' may be included. In some cases, words such as 'cucumber' There is a word consisting of only numeric characters.

In this case, when a numeral character is included in a general word such as 'sha' of 'congratulatory message' or 'sham' of 'ginseng', the 'joe' of 'congratulatory message' If the character is not a numeric character, check the words such as 'congratulatory message' or 'ginseng' as a word indicating the amount of money.

In addition, the word consisting of only numeric characters such as 'cucumber' or 'grilled' has a meaning that a word indicating the unit of currency between the numeral 'o' and the numeral 'i' , Ten, etc.) are absent. Therefore, it is also not possible to judge a word connected with only numeric characters as a word indicating the amount of money in such currency units or digits.

Also, for example, if the voice information is 'Gyeonggi-do 23,000 5,000 won', there is a place to mistake 'Icheon' as the amount. In such a case, the word '30,000 won' It is possible to judge that the number of digits is more than the amount of 'Icheon', which is one digit word, and furthermore, the character of 'Yen' is included in the 30,000 won, It extracts the word '30,000 won' as the amount information rather than the word 'Icheon'.

In addition, for example, when the voice information is 'Gyeonggi-do 2,000 won,' the word 'five thousand' appears after the word 'Icheon', and when these two words are combined, it becomes the same word as 'two thousand five thousand five thousand' The speech analysis unit 30 of the present invention excludes the word " Icheon " from the money item and extracts it.

On the other hand, if the voice information is 'Gyeonggi-do 2,133 won', it is also possible to store the word 'Gyeonggi-do' in advance to prevent the 'Icheon' from being recognized as an amount when the word 'Gyeonggi-do' , And when the voice information is the same as' Sacheon Magnetic Field of 500 Won ', it is judged that the word' Magnetic Field 'is included between the numbers of' Sacheon 'and' And the amount of '100,000 Won' among the '500 Won'.

In addition, when the voice information is input without omitting the amount itself, the user can refer to the category, asset, customer, and type included in the voice information, The amount can be deducted from the input amount, or the amount can be displayed as 0 when the amount itself can not be extracted at all.

6) Extraction of the type item of the transaction

If the word 'import', 'expenditure', or 'transfer' is included in the voice information based on the voice information transmitted from the voice recognition engine 20, Extract the type.

However, if the words 'import', 'expenditure', or 'transfer' are not included, the type of transaction is extracted based on the item of 'category'.

In addition, if the voice information includes words such as 'installment' and 'month' in the extraction process of the type item, it is preferable that the total amount is automatically divided into the number of months.

7) Extracting details of the transaction

The voice analyzing unit 30 analyzes the date information, the time information, the time information, the category information, and the asset information of the corresponding transaction as described above, based on the voice information transmitted from the voice recognition engine 20, , 'Amount', 'customer', 'type', and extracts it as a detail item.

However, if the items such as the account, the account, the date, the time, the amount, and the type are not extracted after extracting the detailed contents as described above, the voice analysis unit 30 searches the past transaction information, It is possible to automatically find the items of the transaction information by searching for transaction information with high degree of matching.

For example, if the past transaction information of the user has stored the contents of "March 10, 2010 cash transportation fee 3,000 won bus fee", if the user simply inputs "bus fare 1,500 won" on March 15, The voice analysis unit 30 may not be able to extract items of 'asset (account)' and 'transportation cost (category)'. At this time, Based on the search result, the contents of the details are inputted as "cash" for the contents to be inputted on March 15, and "transportation fee" for the category.

On the other hand, the transaction information display unit 40 enhances the accuracy of transaction information input by providing a result screen displaying the analyzed transaction information so that the user can confirm whether the information inputted by the user is well converted into transaction information And enhances user convenience.

The storage unit 50 may be a conventional non-volatile memory device or a volatile memory device, and may be a large-capacity random access memory (RAM) or a flash memory device.

3 is a flowchart of an information input method for asset management of the present invention.

The information input method for asset management according to the present invention is implemented in a transaction information input system for a smart device having a speech recognition engine for recognizing a voice uttered by a user and generating voice information and a storage unit for storing information for asset management As an information input method, a category, an asset, a business partner, a date, a time, an amount, a transaction type, and a detail content for transaction information input and asset management are extracted from the voice information inputted by the user, And storing the information in the storage unit.

Specifically, when a user speaks one or more items of a category, an asset, a customer, a date, a time, an amount, a transaction type, and a detailed content of the transaction, Step S 100); A voice information generating step (step S 200) of recognizing the voice inputted by using the voice recognition engine of the smart device and generating voice information composed of the recognized contents; A transaction information extracting step (S 300) of extracting transaction information of the transaction based on the voice information generated by the speech recognition engine and storing the extracted transaction information in the interlocked storage unit; And displaying the extracted transaction information on the display means of the smart device (Step S 400).

The 'category' item means the purpose of the transaction as described above. For example, the categories may include meals, athletics, cosmetics, transportation, hospitals, salaries, insurance, and the like. Typically, a large number of users, such as meals, athletics, As a category, and stores the item in the storage unit.

The category and its items can be directly input by the user and stored and managed.

On the other hand, it is possible to automatically designate or select a category having the latest or latest frequency of use, and in the case of an item that is not in a preset category or does not correspond to a preset category, the content can also be directly input by the user.

In addition, the category may be set as an upper category indicating the purpose of the transaction in a comprehensive manner and a lower category subdivided in the contents of the upper category. When the upper category is 'transportation charge' as shown in FIGS. 2A and 2B Sub categories such as 'Bus', 'Subway', 'Taxi', 'Automobile' can be listed as subcategories of the 'Transportation Expense' category.

The 'assets' of the corresponding transaction are the items in which the user's own goods such as demand account books, periodic savings accounts, bank accounts such as regular accounts, credit cards, check cards, transportation cards, loan accounts, insurance accounts, Lt; / RTI > At this time, the 'asset' item may be set as an upper asset item which comprehensively indicates the contents of the asset and a lower asset item which subdivides the contents of the higher asset.

The 'amount' item of the above transaction indicates the amount of the transaction amount.

The 'customer' item of the transaction indicates the information of the counter party.

At this time, the 'customer' item may be set as an upper customer who comprehensively represents the content of the asset and a lower customer who subdivides the upper customer.

The 'type' item of the corresponding transaction includes an 'income' item in which the transaction is transferred to the user's asset item, an item 'expense' (withdrawal) in which the amount is withdrawn from the asset item, and an item It points to the 'transfer' items to be deposited.

The 'details' of the corresponding transaction may include other transactions that do not correspond to any of the date, time, category, asset, amount, customer, or type of the transaction Point to the item.

The step S 100 of inputting the user's voice to the smart device is performed through a microphone built in the smart device, and recognizes the input voice using the voice recognition engine of the smart device, The step of generating information is performed by the operation of a known speech recognition engine, and a detailed description thereof will be omitted.

The step of extracting the transaction information from the voice information in step S 300 includes a category extraction step S 310, an asset extraction step S 320, a customer extraction step S 330, a date and time extraction step S 340 (S 350), a transaction type extraction step (S 360), and a detail content extraction step (S 370).

4A to 4F are flowcharts of the steps of extracting the transaction information (S 300) of FIG. 3. The category extracting step (S 310) of FIG. 4A stores the words included in the voice information generated by the voice recognition engine Searching (S 311) for matching whether the items of the category previously stored in the contents match with the words of the items stored in the category; If it is determined that a word matching the pre-stored item is included in the voice information (S 312), and if the word corresponding to the pre-stored item is included in the voice information as a result of the determination, (S 313).

For example, as shown in FIG. 5A, if the voice information input by the user is' 5,000 won cash on a meal ', if the word' meal 'is included in the category item, the category of the transaction information is' '.

In this process, when the words included in the voice information coincide with the items corresponding to the items of the sub-category, they are automatically designated as categories corresponding to the upper category of the corresponding sub-category.

For example, as shown in FIG. 5B, if the voice information inputted by the user is 'lunch parcel bargain only 1,000 won cash', there is a word 'lunch' in the words of the item of the subcategory, If the category is included in the item 'meals' as the upper category, the category of the transaction information is automatically designated as 'meals', and the upper and lower categories are displayed in parallel with 'meals / lunches' as shown in FIG. 5B .

On the other hand, if the voice information transmitted in step S 312 does not include a word that matches the categories stored in the storage unit, among the past transaction information stored in the storage unit, at least one word (S 314). If there is a word matching the past transaction information (S 314 '), the designated category is designated as the category item on the past past transaction information (S 315).

For example, in the case where voice information such as 'cash meal only 1,000 won paris baguette' is input while 'Paris baguette' is stored in the past transaction information, the transaction information of 'Paris baguette' , It is possible to extract the 'meal / lunch' category, which is the category of the past transaction information, as shown in FIG. 5B.

However, if any one of the words included in the transmitted voice information does not have a word matching the past transaction information, the category of the corresponding voice information is not specified (S 316) (S 317).

However, the present invention is not limited to this, and depending on the embodiment, the user can automatically select or display the category with the latest or latest frequency of use.

Also, in the case of an item that is not in a preset category or does not correspond to a preset category, the contents of the category may be directly input by the user.

As a method for recommending a category most appropriate to the word, a voice message is transmitted to an external server using a wireless communication module built in the smart device, and then the transaction information of the user, which is stored in the server, If there is a word matching the past transaction information, it is determined whether or not at least one word among the words included in the voice information exists in the transaction information of the user. If there is a word corresponding to the past transaction information, can do.

In addition, the information input method for asset management of the present invention can more easily input transaction information by using a promise word designated by the user.

That is, the user registers the appointment word as the word 'frequently used' through the input window of the embodiment as shown in FIG. 5C, and the category, the asset, the amount, the customer, the transaction type, The user selects one or more of these items, and stores them in the storage unit.

Then, when the user inputs the words of the appointment word as voice information, the items of the category, the asset, the amount, the customer, the transaction type, and the details specified in the promise word are automatically generated and input as transaction information.

For example, when a user designates 'cash (asset) transportation cost (category) 3,000 won (amount) and bus fee (details)' as an appointment word 'bus fee' The transaction information of 'cash (asset) transportation cost (category) 3,000 won (amount) and bus fee (details)' is automatically input as shown in FIG. 5D. The promissory input method can be used to increase the convenience and accuracy of the input when the user frequently uses the transaction.

The asset extraction step (S 320) of FIG. 4B may include searching a word included in the voice information generated by the voice recognition engine to find whether there is a word matching one of the asset items previously stored in the storage unit S 321); If it is determined in step S322 that a word matching the pre-stored asset item is included in the voice information, if the word corresponding to the pre-stored asset item is included in the voice information, (Step S323).

Also, if the words included in the voice information match the items corresponding to the item of the lower asset, the transaction information can be extracted as the asset corresponding to the upper asset of the lower asset.

On the other hand, if the voice information transmitted in step S322 does not include a word matching the stored items of the asset, at least one word among the words included in the voice information among the past transaction information stored in the storage unit (S324). If there is a word matching the past transaction information (S324 '), the designated asset is designated as an asset item on the past transaction information (S325).

However, if any one of the words included in the delivered voice information does not have a word matching the past transaction information, the asset of the corresponding voice information is not specified (S326) (S327).

However, the present invention is not limited to this, and depending on the embodiment, it is possible to automatically designate or display an asset having the latest or latest frequency of use, thereby allowing the user to select it. In addition, if the item of the specified asset does not exist or does not correspond to the preset asset item, the contents of the corresponding asset item can be directly inputted by the user.

The method of recommending the appropriate asset item to a user may include transmitting the voice information to an external server using a wireless communication module built in the smart device as in the case of recommendation of the category, Information on the transaction information of the other user, whether there is at least one word among the words included in the voice information, and if there is a word corresponding to the past transaction information, Recommended.

The step of extracting a customer (S 330) in FIG. 4C may include searching for a word that matches any of the customer items stored in the storage unit among the words included in the voice information generated by the speech recognition engine S 331); If it is determined that a word matching the pre-stored account item is included in the voice information (S332), if the word matching the pre-stored account item is included in the voice information, (S333) of extracting the item as a customer item of the customer.

In addition, when the words included in the voice information match the items corresponding to the items of the lower customer, the transaction information can be extracted as the customer corresponding to the upper customer of the lower customer.

On the other hand, if the voice information transmitted in step S332 does not include a word that matches the items of the account stored previously, at least one word among the words included in the voice information among the past transaction information stored in the storage unit (S334). If there is a word matching the past transaction information (S334 '), the designated customer is designated as a customer item on the past transaction information (S335).

However, if any one of the words included in the transmitted voice information does not have a word that matches the past transaction information, the customer of the voice information is not specified (S336) (S337).

However, the present invention is not limited to this, and according to an embodiment, it is possible to automatically designate or display a customer with the latest or latest frequency of use, thereby allowing the user to select the customer. In addition, if there is no customer or does not correspond to a predetermined customer, the user can directly input the customer item and contents.

As a method for recommending an appropriate account item to a user, the voice information is transmitted to an external server using a wireless communication module built in the smart device, similar to the recommendation of the category and the asset item, If there is a word that matches at least one word among the words included in the voice information in its own transaction information or transaction information of another user, Can be recommended to the user.

In addition, when the customer is extracted, the location information of the customer may be extracted together, and the customer close to the location information of the user at the time of inputting the voice may be automatically extracted.

In addition, if the voice information transmitted in the step S 312 does not include a word matching the categories stored in the voice information transmitted in the step S 312, If the voice information transmitted in step S332 does not include a word matching the pre-stored items of the account, the category information included in the voice information among the past transaction information stored in the storage unit The information of the category or the asset or the customer or the detail in the past transaction information having a large number of matching items and the accuracy is classified into the category or the asset or the details of the customer or the details .

The date and time extracting step (S 340) of FIG. 4 d is a step of extracting the date and time from the voice information generated by the voice recognition engine by using a character representing a number such as a number or a zero, a, a, a, a, a, a, a, (Hereinafter referred to as "numeric character") and a date unit such as year, month, and day are included, the date information is converted into a date and date information of the transaction is extracted (S341).

At this time, if either the year, month, or day of the date information is omitted, the date information is automatically extracted based on the date on which the voice information is input (S342).

For example, when only the date is confirmed from the voice information, the year and the month at the input time of the voice information are automatically extracted. If only the month and date are confirmed from the voice information, If the year, month, and day are not confirmed from the voice information, the voice information is automatically extracted as the date of input of the voice information.

If the voice information includes words (yesterday, yesterday, the day before, the day after tomorrow, the day after tomorrow, the day after tomorrow, and the like), the word indicating the date is converted into a corresponding date and extracted do. That is, if the date is not extracted from the audio information in the same manner as described above, it is determined whether or not other words indicating the date, that is, yesterday, today, tomorrow, yesterday, yesterday, , And compares the extracted date information with the date when the voice information is input, thereby extracting the date information.

4E, if the number is confirmed in the voice information generated by the voice recognition engine and the word of the currency (circle, dollar, yen, etc.) is included in the voice information, the money is extracted as the amount of money (S351) .

In addition, the voice information includes a unit character (e.g., billion, million, thousand, hundred, ten, etc.) indicating the number of digits or digits and the number of digits of the currency but does not include the unit of currency If it is determined that there is a word indicating the denomination of money among the words included in the voice information among the past transaction information stored in the storing unit (S352), and if there is a word indicating the denomination of money in the past transaction information (S353), the money item is extracted in units of the corresponding currency (S354).

As described above, in the unit of the currency used for the search, the stored main use currency is set and used in the storage unit 50. When only the numbers not including the currency unit are extracted, May be combined with a number and extracted as an amount of money item. In some embodiments, one of the denominations of the amount of money in the past transaction information that has been searched may be selected and combined with a number to be extracted as a money item.

In addition, when the voice information is input without omitting the amount itself, the user can refer to the category, asset, customer, and type included in the voice information, The amount can be deducted from the input amount, or the amount can be displayed as 0 when the amount itself can not be extracted at all.

If the money itself is omitted or the money can not be extracted to the transmitted voice information, the voice analysis unit 30 may store at least one of the words included in the voice information among the past transaction information stored in the storage unit 50 And if there is a word that matches the past transaction information, the money may be extracted as a money item on the corresponding past transaction information

The transaction type extraction step (S 360) of FIG. 4F extracts the type corresponding to the included word when the word 'import', 'expenditure', or 'transfer' is included in the voice information generated by the speech recognition engine (S 361).

However, if the words 'import', 'expenditure', or 'transfer' are not included, the type of transaction is extracted based on the past transaction information of the 'category' (S 362). For example, if the voice information input by the user is '5,000 won cash on the dining table surface', it is not clear whether the voice information is an income or expenditure or a transfer, as in the example shown in FIG. 5A, Since the past transaction information is 'expenditure', the transaction type is extracted as 'expenditure'.

If the voice information includes words such as 'installment' and 'month' in the process of extracting the type item, the total amount may be divided by the number of months.

If the voice information includes words such as daily, weekly, monthly, weekly, weekly, and monthly words in the extraction process of the type item, the details may be recognized as regularly repeated contents.

The detailed content extracting step (S 370) is a step of extracting the detailed contents (S 370) from the 'date', 'time', 'category', 'asset', 'amount', ' And extracts other items that are not applicable to one item and extracts them as items of detailed contents.

However, if items such as an account, a customer, a date, a time, an amount, and a type are not extracted after extracting the detailed contents as described above, the past transaction information is searched to find transaction information having high degree of agreement with the detailed information It is possible to automatically extract the items of the corresponding transaction information, which has been described with an example in the detailed content item extraction process of the transaction described above, and thus a detailed description thereof will be omitted.

Meanwhile, the information input method for asset management including each of the above-described steps of the present invention may be implemented as an aspect of a recording medium containing a computer program.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed embodiments, And such modifications, alterations, and substitutions are to be construed as being within the scope of protection of the present invention.

One; Smart device
2; Display means
3; MIC
4; Key input module
10; Voice input unit
20; Speech recognition engine
30; Voice analysis section
40; Transaction information display section
50; The storage unit
100; server

Claims (23)

A transaction information input system of a smart device (1) provided with a display means (2) and a microphone for voice input (3)
A voice input unit 10 for receiving a user voice uttering information for inputting transaction information from the microphone 3 of the smart device 1;
A voice recognition engine 20 for recognizing the voice input to the voice input unit 10 and generating voice information composed of the recognized voice;
A voice analysis unit 30 for extracting transaction information of the transaction based on the voice information generated by the voice recognition engine 20 and storing the extracted transaction information in the interlocked storage unit 50;
A transaction information display unit (40) for displaying the transaction information extracted by the voice analysis unit (30) on the display means (2) of the smart device (1); And,
Wherein the information for inputting transaction information input to the voice input unit (10) is at least one of a category, an asset, a customer, and a detailed content. delete The speech analyzing apparatus according to claim 1, wherein the speech analyzing unit (30)
Searching whether words included in the voice information match any one of the items stored in the storage unit 50 or stored in the category or the asset or the item of the customer,
When a word matching the previously stored category item is included in the voice information, the matching category item is designated as a category,
If the voice information includes a word that matches the stored asset item in the search result, the matching asset item is designated as an asset,
And if a word that matches the pre-stored account item is included in the voice information, the corresponding account item that matches is designated as the account holder. 4. The method according to claim 3, wherein if the transferred voice information does not include a word that matches a category or an item previously stored in an asset or a customer,
The voice analysis unit 30 searches the past transaction information stored in the storage unit 50 to find whether at least one item among the categories or assets included in the voice information,
A transaction information inputting system for a smart device using speech recognition, characterized in that information of a specified category or an asset, a customer, or details is specified on a past transaction information having a large number of matching items and a high accuracy as a category, an asset, . The system according to claim 1, wherein the information for inputting transaction information, which is input to the voice input unit (10)
Date, time, and amount of money in the smart device. 6. The method of claim 5,
The voice analysis unit 30 converts the voice information into a date when the voice information includes numeric or numeric characters and includes a date unit such as year, month, and day, extracts date information of the corresponding transaction,
Wherein when the date, month or date of the date information is omitted, the date on which the voice information is input is extracted as date information. 6. The method of claim 5,
Wherein the voice analysis unit (30) extracts the number of the voice information as a monetary amount when the voice information includes a word of a unit of currency. The system according to claim 1, wherein the information for inputting transaction information, which is input to the voice input unit (10)
A transaction type, and a detailed information. 9. The method of claim 8, wherein the voice analysis unit (30) extracts a transaction type as a corresponding word when the voice information includes the word " import &
If the said income or expenditure or transfer does not include the word,
The method comprising the steps of: detecting whether a word included in the voice information exists in an item of a category stored in the storage unit; and if the word exists, extracting a transaction type of the category as a transaction type of the inputted voice information Transaction information input system. The method according to claim 9, wherein, when the voice information includes words such as 'installment' and 'month' in the extraction process of the type item, the voice analysis unit (30) divides the total amount by the number of months, Extracting the transaction information for the smart device using the voice recognition. A method for inputting information for asset management using a transaction information input system for a smart device having a voice recognition engine for generating voice information and a storage unit for storing information for asset management,
A user voice input step (step S100) in which a user's voice is input to the smart device when the user speaks information of the transaction;
(Step S 200) of generating voice information for recognizing a voice input by using the voice recognition engine of the smart device and generating voice information composed of the recognized contents;
An extraction step (S 300) of extracting transaction information of the transaction based on the voice information generated by the speech recognition engine and storing the extracted transaction information in the interlocked storage unit;
And displaying the extracted transaction information on the display means of the smart device (Step S 400)
The transaction information extracting step S 300 includes at least one of a category extracting step S 310, an asset extracting step S 320, a customer extracting step S 330 and a detail extracting step S 370. Wherein the information is input by a user. delete 12. The method of claim 11,
Searching whether words included in the voice information match any one of the category or the asset or the item of the customer stored in the storage unit,
When a word matching the previously stored category item is included in the voice information, the category item is designated as a category,
If the voice information includes a word that matches the previously stored asset item, the corresponding asset item is designated as an asset,
And when the word corresponding to the stored account item is included in the voice information, the corresponding account item is designated as the account holder. 14. The method of claim 13,
If the delivered voice information does not include words that match the categories or assets previously stored in the voice information, or the items of the customer,
And searching for at least one item of at least one of a category, an asset, a transaction or a detailed content included in the voice information among the past transaction information stored in the storage unit,
Wherein information of a specified category or an asset, a customer or a detail is specified as a category, an asset, a customer, or details in past transaction information having a large number of matching items and high accuracy. 15. The method of claim 14,
If any one of the words included in the delivered voice information does not have a word corresponding to the past transaction information,
And recommends information on a category or an asset or a customer most suitable to the word from the past transaction information stored in the server connected to the smart device to the user. 15. The method of claim 14,
And extracting a customer who is close to the location information of the user at the time of inputting the voice by automatically extracting the location information of the customer when the customer is extracted. 12. The method of claim 11,
An item of a category, an asset, an amount, a customer, a transaction type, and a detail of the promise word and the promise word are specified and stored in the storage unit,
When a word of the promise word is inputted as voice information, items of categories, assets, amounts, customers, transactions, and details specified in the promise word are automatically generated and input as transaction information Information input method. The method as claimed in claim 11, wherein the step (S 300)
(S 340), an amount extracting step (S 350), and a transaction type extracting step (S 360). 19. The method of claim 18, wherein the extracting the date and time (S 340)
If the voice information includes numeric or numeric characters and includes date units such as year, month, and day, converts the converted date into date, extracts the date information of the corresponding transaction,
And when the date, month, or day of the date information is omitted, the date on which the voice information is input is extracted as the date information. 19. The method of claim 18, wherein the extracting step (S 350)
When the number is confirmed in the voice information and the word of the unit of currency is included, it is extracted as the amount,
When only numbers that do not include a currency unit are extracted, the main used currency unit stored in the storage unit is combined with a number and extracted as a money item,
If it is not possible to extract the money from the voice information, it is checked whether at least one word among the words included in the voice information among the past transaction information stored in the storage unit exists. If there is a word matching the past transaction information, And the amount of money is extracted from the past transaction information. 19. The method of claim 18, wherein the transaction type extraction step (S 360)
When the voice information includes the word import, expenditure or transfer, a transaction type of the word is extracted,
If the said income or expenditure or transfer does not include the word,
The method comprising the steps of: detecting whether a word included in the voice information exists in an item of a category stored in the storage unit; and if the word exists, extracting a transaction type of the category as a transaction type of the inputted voice information . 22. The method of claim 21, wherein if the item of the category does not exist,
Searching whether there is at least one word among the words included in the voice information among past transaction information stored in the storage unit,
And if there is a word corresponding to the past transaction information, the transaction type is extracted on the past transaction information. A recording medium on which a program to be executed including the information inputting method for asset management according to any one of claims 11 to 13 is recorded.

Images