KR20230127530A - A Recommendation Method For Music Score And Concerto Gathering And Device Thereof - Google Patents

Abstract

This relates to a method and recommendation device for recommending scores and concertos. The method for recommending scores and concertos is based on attribute values of atmosphere, speed, and difficulty for a song of interest among user questions and answers, and attribute values for atmosphere, speed, and difficulty among user questions and answers. Calculating a user attribute vector Converts pre-stored information about difficulty, speed, and genre of the score into attribute values for each attribute of the information, analyzes the musical score, and plays based on the pre-stored information about the musical instrument's pitch range Musical instrument property values are generated, key signatures are calculated based on the key signature of the score, and atmosphere property values are generated based on the calculated key signatures, and sheet music property vectors are converted into property values for difficulty, speed, genre, playing instrument, and atmosphere properties. Generating a sheet music property vector for each sheet music for a plurality of sheet music stored in a storage unit, generating a list of recommended sheet music in order of high similarity based on the similarity between the user property vector and the plurality of sheet music property vectors; and and generating a recommended concerto list based on playing instruments of the recommended score list and playing instrument information of the user property vector.

Description

Sheet Music and Concerto Recommendation Method and Recommendation Device {A Recommendation Method For Music Score And Concerto Gathering And Device Thereof}

A method and apparatus for recommending sheet music and concerto based on user properties and sheet music properties.

A musical score and concerto recommendation device is a system that analyzes a user's tendency and recommends content suitable for the analyzed user's tendency. Due to the recent activation of content providing services, various user-customized content recommendation methods are popularized in daily life.

In addition, with the diversification of users' hobbies, the number of users who play musical instruments as a hobby is increasing, but those who play musical instruments for the first time and learn by themselves select scores based on non-quantified information from Internet media or select scores. Even if it is, there is a problem of selecting a score that is not desired by the user or has a difficult difficulty level. In addition, a service providing sheet music is lacking in a device that recommends a suitable sheet music to a user by performing a search with limited information on a song title and composer.

In addition, even if a user-customized score is selected, in the case of a score that requires concerto, it is difficult to recruit concertos with other users suitable for the score.

Republic of Korea Patent Registration No. 10-1057919 ("Customized music recommendation method through analysis of users' playlist", Sungkyunkwan University Industry-University Cooperation Foundation, 20161.08.11)

It is possible to provide a score and concerto recommendation method and a recommendation device for recommending a score and concerto suitable for a user's individual characteristics by matching properties of a plurality of sheet music and user's properties.

An embodiment of a method for recommending sheet music and concerto includes calculating a user attribute vector based on attribute values of atmosphere, speed, and difficulty for a music of interest among user question answers and attribute values for mood, speed, and difficulty among user question answers. converting information about difficulty, speed, and genre stored in advance into attribute values for attributes of each information, analyzing the musical score and generating attribute values of playing instruments based on pre-stored information about the musical instruments, Generating a score property vector with attribute values for difficulty, speed, genre, playing instrument, and mood by calculating a key based on the key signature of the score and generating attribute values of atmosphere based on the calculated key; Generating a plurality of sheet music stored in a storage unit for each sheet music, generating a recommended sheet music list in order of high similarity based on the similarity between the user property vector and the plurality of sheet music property vectors, and playing instruments in the recommended sheet music list, and A step of generating a recommended concerto list based on playing instrument information of the user attribute vector may be included.

In addition, the step of calculating the user attribute vector according to an embodiment may include calculating an average value of attribute values for the music of interest and attribute values included in the user's Q&A for each attribute such as atmosphere, speed, and difficulty, and then calculating the average of the attribute values for the user per attribute. It may be characterized by calculating an attribute value.

In addition, in the step of calculating the sheet music attribute vector according to an embodiment, if the calculated key is major, the atmosphere is determined to be bright to generate the property value of the atmosphere, and if the calculated key is minor, the atmosphere is determined. It may be characterized in that the attribute value of the atmosphere is generated by determining the darkness.

Also, in the step of calculating the score property vector according to an embodiment, the pre-stored level of difficulty of the score may be stored based on user evaluation of the score.

In addition, the step of calculating the score attribute vector according to an embodiment may be characterized in that the attribute value of the speed of the score is determined as a value on a lookup table stored in a storage unit corresponding to the speed of the score.

In addition, in the step of calculating the score attribute vector according to an embodiment, the pitch range of the score is generated and analyzed in the form of a vector including an octave attribute and a term name attribute based on a lookup table stored in a storage unit. can be done with

In addition, the step of generating a recommended sheet music list according to an embodiment may include selecting a plurality of sheet music property vectors including performance instrument and genre property values of a user property vector among a plurality of sheet music property vectors, and selecting a plurality of sheet music property vectors selected. It may be characterized in that the similarity between the vector and the numerical values of the atmosphere, speed, and difficulty properties of the user property vector is calculated, and a score corresponding to the score property vector having a high degree of similarity is given priority to generate a list of recommended music scores.

In addition, according to an embodiment, after the step of calculating the score property vector, it is checked whether information about the code is included in the score, and if there is no information about the code, the triad code is obtained by using the notes included in the bar as input. It may be characterized by further comprising the step of determining the code of the musical score with a chord prediction model as a label.

In addition, according to an embodiment, triad codes used as labels may be characterized in that they include four major codes, minor codes, augmented codes, and diminished codes.

It is possible to recommend scores and concertos suitable for individual characteristics of users based on the score property vectors for a plurality of scores and the user property vectors for user queries and pre-stored user information.

1 is a block diagram of a sheet music and concerto recommendation apparatus according to an embodiment.
2 is a block diagram of a sheet music analysis unit according to an embodiment.
3 illustrates an example of generating a score property vector from a score according to an embodiment.
4 illustrates a concept of calculating a vector attribute value for a musical score range in a musical score according to an embodiment.
5 is a lookup table for calculating playable musical instruments based on musical score ranges according to an embodiment.
6 is a lookup table for calculating a key based on a key signature in a musical score according to an embodiment.
7 is a lookup table for calculating a velocity vector attribute value in a sheet music according to an embodiment.
8 illustrates a concept of determining a code of a musical score according to an embodiment.
9 illustrates an example of a user query according to an embodiment.
10 illustrates an example of calculating a user attribute vector based on a user query answer according to an embodiment.
11 illustrates a concept of recommending sheet music based on a user property vector and a sheet music property vector according to an embodiment.
12 is a flowchart of a method for recommending sheet music and concerto according to an embodiment.

Advantages and features of the present invention, and methods of achieving them, will become clear with reference to the embodiments described below in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but may be implemented in various different forms, and only these embodiments make the disclosure of the present invention complete, and common knowledge in the art to which the present invention belongs. It is provided to completely inform the person who has the scope of the invention, and the present invention is only defined by the scope of the claims.

The terms used in this specification will be briefly described, and the present invention will be described in detail.

The terms used in the present invention have been selected from general terms that are currently widely used as much as possible while considering the functions in the present invention, but these may vary depending on the intention of a person skilled in the art or precedent, the emergence of new technologies, and the like. In addition, in a specific case, there is also a term arbitrarily selected by the applicant, and in this case, the meaning will be described in detail in the description of the invention. Therefore, the term used in the present invention should be defined based on the meaning of the term and the overall content of the present invention, not simply the name of the term.

When it is said that a certain part "includes" a certain component throughout the specification, it means that it may further include other components without excluding other components unless otherwise stated. In addition, terms such as “unit”, “module”, and “unit” used in the specification mean a unit that processes at least one function or operation, and includes software, hardware components such as FPGA or ASIC, or software and hardware. It can be implemented as a combination of However, terms such as “unit”, “module”, and “unit” are not meant to be limited to software or hardware. A “unit”, “module”, “unit”, etc. may be configured to be in an addressable storage medium or configured to reproduce one or more processors. Thus, as an example, terms such as ""unit", "module", and "unit" refer to components such as software components, object-oriented software components, class components, and task components, and processes. fields, functions, properties, procedures, subroutines, segments of program code, drivers, firmware, microcode, circuits, data, databases, data structures, tables, arrays and variables. .

Hereinafter, with reference to the accompanying drawings, embodiments of the present invention will be described in detail so that those skilled in the art can easily carry out the present invention. And in order to clearly explain the present invention in the drawings, parts irrelevant to the description are omitted.

Terms including ordinal numbers such as “first” and “second” may be used to describe various elements, but elements are not limited by the terms. Terms are only used to distinguish one component from another. For example, a first element may be termed a second element, and similarly, a second element may be termed a first element, without departing from the scope of the present invention. The term “and/or” includes a combination of a plurality of related items or any one of a plurality of related items.

Hereinafter, with reference to the accompanying drawings, an embodiment of a score and concerto recommendation method based on user properties and sheet music properties and a sheet music and concerto recommendation device 1 based on user properties and sheet music properties will be described.

1 shows a block diagram of a sheet music and concerto recommendation device 1 .

The sheet music and concerto recommendation apparatus 1 generates a user property vector based on a user query and pre-stored user information, generates a sheet music property vector for a plurality of sheet music, and recommends a sheet music having a sheet music property vector similar to the user property vector. And, based on the playable instruments included in the recommended score, a concerto may be recommended to the user. In addition, the sheet music and concerto recommendation device 1 may include an input unit 100 , a storage unit 200 , an output unit 300 and a processor 400 .

The input unit 100 includes code information for generating the code prediction model 443 of the score and concerto recommendation device 1, a plurality of scores for generating a score property vector, information on the difficulty and genre of the score, and information on queries. Receives the user's answer, the user's location information for recommending a concerto, the time available for concerto, the code for the score, the genre, the atmosphere, the speed, the range, the instrument, the key signature, and the key, and stores the received signal ( 200), the output unit 300, and the processor 400. In addition, the input unit 100 includes at least one command for starting the operation of the sheet music and concerto recommendation device 1 (eg, a command for driving a program for recommending a place (which can be referred to as an app, application, or software). etc.) may be received and transmitted to the processor 400. Depending on the embodiment, the input unit 100 may receive the above-described data when a user physically manipulates a keyboard or a touch screen, or another device (not shown) transmits the above-described data through a wired or wireless communication network. . If necessary, the input unit 100 may transfer received data to the storage unit 200, and in this case, the storage unit 200 may temporarily or non-temporarily store these data. The input unit 100 may include, for example, a keyboard, a mouse, a tablet, a touch screen, a touch pad, a track ball, a track pad, a scanner device, an image capturing module, an ultrasonic scanner, a motion detection sensor, a vibration sensor, a light receiving sensor, a pressure sensor, A proximity sensor and/or a data input/output terminal may be included, but is not limited thereto.

The storage unit 200 stores data input through the input unit 100, required in the processing of the processor 400, generated in the processing of the processor 400, or data according to processing results of the processor 400 temporarily or can be temporarily stored. For example, the storage unit 200 includes a user query, preference information and concerto preference information about properties of musical scores stored in advance by the user, location information of the user, time available for concertos, a plurality of musical scores, attribute information about musical scores, and code prediction. Triad code information for generating the model 443 may be included. The storage unit 200 may include at least one of a main memory device and an auxiliary memory device. The main memory device may be implemented using a semiconductor storage medium such as ROM and/or RAM, and the auxiliary memory device may be a flash memory device (solid state drive (SSD)). ), etc.), SD (Secure Digital) card, hard disk drive (HDD, Hard Disc Drive), compact disk, DVD (DVD) or laser disk, etc. can be implemented based on a device that can store data permanently or semi-permanently. .

The output unit 300 may provide one or more recommended musical scores to the user, and further provide the user with information about the recommended musical scores or information about the instrument list and recommended concerto list included in the recommended musical scores, if necessary. It is also possible to do Also, the output unit 300 may simultaneously or sequentially output at least one query prepared to receive user information or sheet music information. The output unit 300 may provide the above-described information to a user in a visual or auditory form, or may transmit the above-described information to another device through a wired or wireless communication network. Depending on the embodiment, the output unit 300 may include a display, a printer device, a speaker device, an image output terminal, a data input/output terminal, or a communication module.

The processor 400 may generate a query for obtaining user information from a user, or may generate a user attribute vector using information previously or currently input by one or more users. In addition, the processor 400 analyzes a plurality of musical scores stored in the database 210 from the storage unit 200 and analyzes attribute values for each musical score based on information included in each musical score to generate a musical score property vector. In addition, the chord of the score may be determined by generating the chord prediction model 443 based on the input value and the chord label. In addition, the processor 400 may determine a score having a score property vector similar to the user property vector by comparing the individual score property vectors of the plurality of scores with the user property vector.

The processor 400 may execute a program stored in the storage unit 200 in order to execute at least one operation described above. The processor 400 may include, for example, a central processing unit (CPU), a graphic processing unit (GPU), a micro controller unit (MCU), an application processor (AP), ), an electronic control unit (ECU), a micro processor (Micom), and/or at least one electronic device capable of performing other various calculation and control processes. These processing or control devices may be implemented using, for example, one or more semiconductor chips, circuits, or related components alone or in combination.

The processor 400 may include an analysis unit 410 and a recommendation unit 470 .

The analysis unit 410 may receive information from the input unit 100 and the storage unit 200, and based on the received information, generate a user property vector for a user and a sheet music property vector for each sheet music for a plurality of sheet music. . The analyzer 410 may include a query analyzer 420 that generates a user attribute vector and a score analyzer 430 that generates a score attribute vector.

Hereinafter, the query analyzer 420 will be described with reference to FIGS. 9 and 10 .

The query analyzer 420 may receive an answer to the user's query from the input unit 100 and generate attribute values for the user's propensity.

As shown in FIG. 9, the query is for grasping the user's preference for sheet music and musical instruments to be played. Genre, mood, tempo, and difficulty tendency information extracted based on the song), question c asked about the user's preferred atmosphere, question d asked about the user's preferred song speed, and question e asked about the user's preferred performance It can consist of queries about difficulty.

The query analyzer 420 may generate a user attribute vector value by analyzing an answer to the user's query. Specifically, as shown in FIG. 10 , the query analyzer 420 recognizes that the user's preferred musical instrument, which is question a in the question answer content, is the piano, and determines the item for the playing instrument among the user attribute vector values to be the piano. . In addition, the query analyzer 420 selects the most-counted genre among a plurality of songs of interest selected by the user based on the genres of the songs of interest stored in the storage unit 200 based on the answer to question b, which is the user's songs of interest. It can be determined as an item for genre among attribute vector values. That is, if the genre of the three songs selected by the user is counted 2 times for jazz and 1 time for classical music, the query analysis unit 420 determines jazz as the genre item among the user attribute vector values. can

In addition, the query analyzer 420 stores pre-measured scores (attribute values) of selected items for atmosphere (c item), speed (d item), and difficulty (e item) among items selected by the user through a query. Based on the attribute values for the mood, speed, and difficulty of the music of interest selected by the user stored in 200, values for each item of mood, speed, and difficulty among user attribute vector values may be determined. For example, the query analyzer 420 determines the mood among the items of the user attribute vector value as 1 point for the user's answer to question c, and the mood score for each song selected by the user as a favorite song (1 point, 1 point) , 1 point) is received from the storage unit 200 and an average value thereof is calculated to determine the attribute vector value for the atmosphere as “1”. In addition, the query analysis unit 420 determines the speed among the items of the user attribute vector value as 7 points for the user's answer to item d, and the speed score for each song (3 points, 1 point, 5 points) of the song selected by the user as a favorite song. point) is received from the storage unit 200 and an average value thereof is calculated to determine the attribute vector value for the atmosphere as “4”. In addition, the query analysis unit 420 determines the difficulty level of the user attribute vector value items as 5 points for the user's answer to question e, and determines the difficulty score for each song of the user's favorite song (2 points, 1 point, 4 points). point) is received from the storage unit 200 and an average value thereof is calculated to determine the attribute vector value for the atmosphere as “3”.

The query analyzer 420 may generate a user attribute vector based on pre-stored information about a user attribute vector attribute item and a song of interest among answers to the user's query.

In addition, the query analyzer 420 additionally includes the user's sheet music preference tendency or the user's location information, preferred number of concertos, and time available for concerto input through the input unit 100 or stored in the storage unit 200 in the user attribute vector You can create a user attribute vector.

Hereinafter, the sheet music analyzer 430 will be described with reference to FIGS. 2 to 8 .

The sheet music analysis unit 430 may analyze information on a plurality of input musical scores to determine a code for each of a plurality of musical scores or generate a musical score property code. The sheet music analyzer 430 may include a code generator 440, a property analyzer 450, and a sheet music property vector generator 460.

The code generation unit 440 may determine the code of the score to be analyzed. The code generation unit 440 detects whether or not a code is included in the score to be analyzed, and if the code is included, determines the included code as a code of the score, and if the code is not included, the code through the code prediction model 443 can decide The code generation unit 440 may include a code detection unit 441 , a code determination unit 442 and a code prediction model 443 .

The code detection unit 441 may detect whether or not a code is included in the score to be analyzed, and transmit it to the code determination unit 442 .

The code determination unit 442 may determine and output the included code as the code of each bar when the code is included in the score in the code detection unit 441 . In addition, the code determination unit 442 may determine the code predicted through the code prediction model 443 as the code of each measure when the code is not included in the score in the code detection unit 441 .

The code prediction model 443 may predict the code for each bar by inputting the constituent sounds included in each bar of the bar in the sheet music when the code detector 441 does not include the code in the sheet music. Specifically, the code prediction model 443 may be designed based on a deep neural network (DNN) model of machine learning. The code prediction model 443 can predict the code for each measure of the score to be analyzed by setting the input value to a component sound included in one measure of the score to be analyzed and setting the label to a Traid (third chord) code. In this case, the input constitutive sounds are changed into numbers and inputted, and the number conversion for the constitutive sounds may be the same as the numerical representation of each field name shown at the bottom of FIG. 4. In addition, the code prediction model 443 can classify the Traid code as a label into only four codes (Major code, Minor code, Augmented code, and Diminished code) as shown in FIG. 8 .

The attribute analyzer 450 may analyze attribute values of attribute vectors included in the score to be analyzed and convert them into text or numerical values. Specifically, the attribute analyzer 450 analyzes the score shown in FIG. 2, recognizes the lowest note as 2 octave wave# (F#) and the highest note as 3 octave sole (G), and sets the range to 2 octave wave# ( It can be determined that it is from F#) to 3 octave Sol (G), and the key signature is ?? It can be determined that it is a major key (D major), and the speed of the song can be determined to be 100 to 108 BPM. The attribute analyzer 450 may include a voice range analyzer 451, a velocity analyzer 452, a key signature analyzer 453, and a tonality analyzer 454.

The voice range analyzer 451 may analyze the range by detecting the highest and lowest notes of the sheet music. For example, in the score shown in FIG. 3, the sound range analysis unit 451 detects that the highest pitch is 3 octave sol (G) and the lowest pitch is 2 octave wave # (F#), and the pitch range is 2 octave wave # (F#). It can be determined that it is up to 3 octaves sol (G).

The voice range analysis unit 451 may convert the name of the highest pitch, the lowest pitch, and the name of the pitch into a vector of attribute values in the form of numbers. Specifically, as shown in FIG. 4 , the sound range analysis unit 451 may convert the name of a musical score into a number, and convert the notes included in the musical score into a vector consisting of an octave and the name of a musical score. For example, if the sound to be converted into a vector is 1 octave degree (C), the sound range analysis unit 451 may convert the vector to v = {1, 1}, and if it is -1 octave (D) You can convert a vector to v={-1, 2}.

The speed analyzer 452 may recognize the speed of a song included in the sheet music and convert it into a numerical value. The speed analysis unit 452 recognizes whether the speed of the song included in the score is written in text (tempo signature, tempo symbol) or BPM numerical form, and converts it into an attribute value corresponding to the recognized tempo. . As shown in FIG. 7 , the speed analyzer 452 may convert the speed of the music in step 10 into an attribute value on the lookup table.

The key signature analyzer 453 may analyze the key signature included in the musical score and use it to calculate the mood attribute value of the musical score through tonal analysis.

The tone analyzer 454 may analyze the composition of the musical score based on the key signature and ending notes analyzed by the key signature analyzer 453 .

Specifically, when the key signature analyzed by the key signature analysis unit 453 is # (sharp), the tonality analyzer 454 is a sound one whole note above the last # (note A) and a note below a minor third (note B). ) becomes a candidate, and A note becomes Major (major key) and B note becomes Minor (minor key). In addition, if the key signature analyzed by the key signature analysis unit 453 is ♭ (flat), the tonality analysis unit 454 is a candidate for the second ♭ note (A note) and a minor third lower (B note). The note A becomes Major (major) and the note B becomes Minor (minor), and among the two candidates, the same note as the end note of the music score can be selected as the corresponding key. For example, the tonality analyzer 454 may determine the tonality corresponding to the musical score based on key signatures and endings, as shown in the lookup table shown in FIG. 6 .

The sheet music property vector generation unit 460 calculates property values for each of a plurality of properties (playing instrument, genre, atmosphere, speed, difficulty) of the sheet music to be analyzed based on the range, speed, key signature, and tonality analyzed by the analysis unit, You can create a sheet music property vector.

Specifically, the score property vector generator 460 compares the pitch range of the score determined by the musical score analysis unit 451 with the pitch range for each instrument stored in the storage unit 200 shown in FIG. By selecting it, the playing instrument can be created as text as the property value of the playing instrument of the score property vector. Also, the score attribute vector generation unit 460 may generate information on the genre of a song included in the score as an attribute value of the genre as a text. In addition, the score attribute vector generation unit 460 sets the attribute value to 1 as bright when the key determined by the composition analysis unit 454 is major, and sets the attribute value to 0 as dark when it is minor. can be set In addition, the score attribute vector generator 460 may set a speed vector value on the lookup table as an attribute value, as shown in FIG. 7 , based on the speed of the score analyzed by the speed analyzer 452 . In addition, the score property vector generation unit 460 converts the difficulty of the song included in the score into an attribute value, or digitizes the opinions of a plurality of other users based on the opinions of other users who have played in advance, averages them, and You can set the difficulty of the song and convert it into an attribute value.

The recommendation unit 470 may determine musical scores and concertos to be recommended to the user based on the user property vector and the sheet music property vector. The recommender 470 includes a score recommender 480 that determines a score to be recommended and a concerto recommender 490 that determines a recommended concerto meeting with other users for playable instruments included in the score based on the determined score. can include

The sheet music recommendation unit 480 may generate a sheet music recommendation list by comparing the user property vector with each sheet music property vector for a plurality of sheet music and selecting a sheet music property vector most similar to the user property vector.

Specifically, as shown in FIG. 11 , the sheet music recommendation unit 480 may calculate a similarity between a user property vector and a plurality of sheet music property vectors. As shown in the first region of FIG. 11, the music score recommendation unit 480 checks whether the corresponding attribute value of the user attribute vector is included in the attribute value of the score attribute vector for the attributes of the musical instrument and genre in which the attribute value of the attribute vector is composed of text. The degree of similarity can be calculated. In addition, as shown in the second region of FIG. 11, the score recommendation unit 480 may calculate the similarity through Equation 1 below for the attributes of the atmosphere, speed, and difficulty in which the attribute values of the attribute vector are composed of numbers.

(A: user attribute vector, B: score attribute vector, i: attribute sequence of attribute vector)

The concerto recommendation unit 490 may recommend a concerto for the recommended sheet music based on the sheet music recommended by the sheet music recommendation unit 480 . Specifically, the concerto recommendation unit 490 may recommend a concerto, which is a group to play the recommended sheet music, based on the user's region information included in the user attribute vector, age, and playing instrument level. In addition, the concerto recommendation unit 490 may display some of the recommended concerto lists based on whether the recommended concerto is performed online, offline concerto region, or age.

Hereinafter, with reference to FIG. 12, a method for recommending sheet music and concerto will be described.

The sheet music and concerto recommendation device 1 receives a user question answer including information about the playing instrument, the music of interest, the preferred mood, the preferred speed, and the preferred level of difficulty in response to the query from the user. A user attribute vector may be calculated based on attribute values of speed and difficulty and attribute values for atmosphere, speed, and difficulty among the user question and answer (S10).

The score and concerto recommendation apparatus 1 converts pre-stored information about difficulty, speed, and genre of a score into attribute values in order to generate a score attribute vector for a plurality of scores (S20), and analyzes and stores the musical score range. Based on the range information of musical instruments stored in advance stored in the unit 200, an instrument corresponding to the range of the score is selected, and the corresponding instrument is designated as the property value of the performance instrument to generate the property value of the performance instrument (S30). can In addition, the score and concerto recommendation apparatus 1 may calculate the key table of the score, calculate the composition of the score based on the calculated key table, and generate an attribute value of the mood based on the calculated key table (S40).

The sheet music and concerto recommendation apparatus 1 may generate sheet music property vectors for each sheet music for a plurality of sheet music based on property values of difficulty, speed, genre, playing instrument, and atmosphere (S50).

The sheet music and concerto recommendation apparatus 1 calculates the similarity between the user property vector and a plurality of sheet music property vectors, and calculates the musical instrument and genre to be played by text counting, and the atmosphere, speed, and difficulty to calculate the numerical similarity, and recommended sheet music in the order of high similarity. A list can be created (S60).

The sheet music and concerto recommendation apparatus 1 may generate a recommended concerto list based on the information of the playing instruments of the recommended music score list and the playing instruments of the user property vector (S70).

Those skilled in the art related to the embodiments of the present invention will be able to understand that it may be implemented in a modified form without departing from the essential characteristics of the above description. Therefore, the disclosed methods are to be considered in an illustrative rather than a limiting sense. The scope of the present invention is shown in the claims rather than the detailed description of the invention, and all differences within the equivalent range should be construed as being included in the scope of the present invention.

1: Sheet music and concerto recommendation device
100: input unit
200: storage unit
300: output unit
400: processor
410: analysis unit
420: query analysis unit
430: sheet music analysis unit
470: recommendation unit
480: sheet music recommendation unit
490: ensemble recommendation unit

Claims (9)

Calculating a user attribute vector based on attribute values of mood, speed, and difficulty for a song of interest among user questions and answers, and attribute values for mood, speed, and difficulty among the user questions and answers;
Pre-stored information on the difficulty, speed, and genre of the score is converted into attribute values for the properties of each information, and the musical range of the score is analyzed to generate the attribute value of the playing instrument based on the pre-stored information on the musical score of the musical instrument. and calculates a tone based on the key table of the score, generates attribute values of mood based on the calculated key, and converts a score attribute vector into attribute values for the difficulty, speed, genre, playing instrument, and mood attributes. generating;
generating the sheet music property vector for each sheet music for a plurality of sheet music stored in a storage unit;
generating a list of recommended musical scores in descending order of similarity based on similarities between the user property vector and the plurality of sheet music property vectors; and
generating a recommended concerto list based on information on playing instruments of the recommended musical score list and playing instruments of the user property vector;
A method for recommending sheet music and concertos that include. According to claim 1,
The step of calculating the user attribute vector,
A user attribute value for each attribute is calculated by calculating an average value of attribute values for the song of interest and attribute values included in the user's question and answer for each of the attributes of the atmosphere, the speed, and the difficulty, and How to recommend a concerto. According to claim 1,
In the step of calculating the score property vector,
If the calculated composition is major, the atmosphere is determined to be bright to generate attribute values of the atmosphere;
If the calculated composition is Minor, the atmosphere is determined to be dark and an attribute value of the atmosphere is generated. According to claim 1,
In the step of calculating the score property vector,
The score and concerto recommendation method, characterized in that the pre-stored difficulty of the score is stored based on user evaluation of the score. According to claim 1,
In the step of calculating the score property vector,
The score and concerto recommendation method characterized in that the attribute value of the speed of the score is determined as a value on a lookup table stored in a storage unit corresponding to the speed of the score. According to claim 1,
In the step of calculating the score property vector,
The vocal range of the score is generated and analyzed in the form of a vector including the octave attribute and the instrument name attribute based on a lookup table stored in a storage unit for octave and instrument name. According to claim 1,
The step of generating the recommended sheet music list,
Among the plurality of score property vectors, a plurality of score property vectors including the performance instrument and genre attribute values of the user property vector are selected, and the atmosphere of the selected plurality of score property vectors and the user property vector is selected. A method for recommending sheet music and concerto, characterized in that by calculating the similarity between the numerical values of the speed and the difficulty attribute, and generating a list of recommended sheet music by prioritizing the sheet music corresponding to the sheet music property vector having a high similarity. According to claim 1,
After calculating the score property vector,
Checking whether the score includes information about the code, and if there is no information about the code, determining the code of the score with a code prediction model using the notes included in the measure as inputs and using the triad code as a label A method for recommending sheet music and concerto, characterized in that it further comprises a step. According to claim 8,
The score and concerto recommendation method, characterized in that the triad codes as the labels are four major codes, minor codes, augmented codes, and diminished codes.