KR20240135943A - Method for recommending support project and apparatus thereof - Google Patents

Abstract

An embodiment of the present invention relates to a method for allowing a device to recommend a support project. The method comprises: a step for obtaining support project information; a step for extracting a hashtag from the support project information; a step for calculating similarity between the support projects based on the hashtag; a step for generating cluster information by clustering the support projects based on the similarity; a step for obtaining user information for a plurality of users and preferred support project information for a support project preferred by the plurality of users; a step for deriving recommended support project information for one user among the plurality of users from the user information and the preferred support project information through a recommended model using a collaborative filtering based algorithm; a step for adjusting the derived recommended support project information by reflecting at least one among the support project information, the hashtag, the similarity, and the cluster information; and a step for providing the adjusted recommended support project information to the one user.

Description

METHOD FOR RECOMMENDING SUPPORT PROJECT AND APPARATUS THEREOF

The present invention relates to a method and device for recommending a support project.

The scale of the project to support small and medium-sized enterprises through the Ministry of SMEs and Startups in our country reached 16.8 trillion won as of 2021. There are currently 3.73 million small and medium-sized enterprises in our country. However, not all small and medium-sized enterprises are receiving support projects. This may mean that some companies may receive benefits in duplicate.

For example, from 2011 to 2013, a total of 299 companies benefited from startup support projects more than four times, and among them, 22 companies benefited more than 65 times. In addition, one company received support of 94.3 billion won through eight projects from five ministries over five years from 2010 to 2014, and another company received support of 30 billion won through 31 projects from five ministries during the same period. From the perspective of public institutions executing support projects, this concentration of support projects on specific companies is a major problem.

On the other hand, there are also difficulties from the perspective of small and medium-sized enterprises. For example, the problems that small and medium-sized enterprises choose are that it is difficult to identify suitable support projects, that they do not know what support projects are available, that they find out about support projects late, and that it is difficult to fill out application documents. In other words, enterprises are having difficulties in searching for and applying for support projects. In addition, when analyzing the perception of government support projects, the results were 'Please tell me, how, I do not know'. This can be summarized as 'Small and medium-sized enterprises do not know how to support support projects, so please tell me'.

Accordingly, the inventor of the present invention completed the present invention after extensive research to develop a service that recommends and provides support projects that meet the needs of users.

Against this backdrop, one purpose of the present embodiment is to provide a support business recommendation service that obtains user information on a plurality of users and preferred support business information on support businesses preferred by the plurality of users, and derives and provides recommended support business information on a support business preferred by one of the plurality of users from the user information and the preferred support business information through a recommendation model using a collaborative filtering-based algorithm.

Meanwhile, other unspecified purposes of the present invention will be additionally considered within a range that can be easily inferred from the following detailed description and its effects.

In order to achieve the above-described object, one embodiment provides a method for recommending a support project, comprising: a step of obtaining support project information; a step of extracting a hashtag from the support project information; a step of calculating a similarity between the support projects from the hashtag; a step of generating cluster information by clustering the support projects based on the similarity; a step of obtaining user information on a plurality of users and preferred support project information on support projects preferred by the plurality of users; a step of deriving recommended support project information for recommendation to one user among the plurality of users from the user information and the preferred support project information through a recommendation model using a collaborative filtering-based algorithm; a step of adjusting the derived recommended support project information by reflecting at least one of the support project information, the hashtag, the similarity, and the cluster information; and a step of providing the adjusted recommended support project information to the one user.

In the above method, the step of calculating the similarity may include converting the hashtag into a vector and calculating the similarity between the hashtags based on the converted vector.

In the above method, the preferred support business information may include information on a number of support businesses preferred by the plurality of users over a certain period of time.

In the above method, the step of deriving the recommended support business information through the recommended model can derive the recommended support business information for each of a plurality of support businesses preferred by the user during the predetermined period.

In the above method, the step of adjusting the derived recommended support project information may exclude some of the support projects included in the derived recommended support project information based on at least one of the support project information, the hashtag, the similarity, and the cluster information.

Another embodiment includes a processor; and a memory storing instructions executable by the processor, wherein the processor, by executing the instructions, obtains support project information, extracts a hashtag from the support project information, calculates a similarity between the support projects from the hashtag, clusters the support projects based on the similarity to generate cluster information, obtains user information about a plurality of users and preferred support project information about support projects preferred by the plurality of users, derives recommended support project information for recommending to one of the plurality of users from the user information and the preferred support project information through a recommendation model using a collaborative filtering-based algorithm, and adjusts the derived recommended support project information by reflecting at least one of the support project information, the hashtag, the similarity, and the cluster information, and provides the adjusted recommended support project information to the one user.

As described above, according to this embodiment, by deriving and recommending a support project that matches the characteristics of the user among the support projects currently being contested, the usability of the support project can be increased and the needs of the user can be met.

And, according to this embodiment, the accessibility of the support project to the users can be increased by satisfying the needs of the users for the support project.

Meanwhile, it is added that even if an effect is not explicitly mentioned herein, the effect and its provisional effect described in the following specification expected by the technical features of the present invention are treated as described in the specification of the present invention.

Figure 1 is a configuration diagram of a system for recommending a support project according to one embodiment.
FIG. 2 is a flowchart illustrating an operation of a device recommending a support project according to one embodiment.
FIG. 3 is a diagram for explaining a method for recommending a support project according to one embodiment.
Figure 4 is a configuration diagram of a device that recommends a support project according to one embodiment.
It is to be understood that the attached drawings are provided for reference only to help understand the technical concept of the present invention, and the scope of the rights of the present invention is not limited thereby.

In describing the present invention, if it is judged that the detailed description of related known functions that are obvious to those skilled in the art and may unnecessarily obscure the gist of the present invention, the detailed description will be omitted.

The terminology used in this application is only used to describe specific embodiments and is not intended to limit the present invention. The singular expression includes the plural expression unless the context clearly indicates otherwise. In this application, it should be understood that the terms "comprises" or "has" and the like are intended to specify the presence of a feature, number, step, operation, component, part or combination thereof described in the specification, but do not exclude in advance the possibility of the presence or addition of one or more other features, numbers, steps, operations, components, parts or combinations thereof.

Hereinafter, embodiments of the present invention will be described in detail with reference to the attached drawings. When describing with reference to the attached drawings, identical or corresponding components are assigned the same drawing numbers and redundant descriptions thereof are omitted.

Figure 1 is a configuration diagram of a system for recommending a support project according to one embodiment.

Referring to FIG. 1, a system (10, hereinafter referred to as “system”) for recommending a support project according to one embodiment may include a device (100, hereinafter referred to as “device”) for recommending a support project, a plurality of user terminals (200-1, 200-2, ..., 200-N), and a database (300).

A device (100), a plurality of user terminals (200-1, 200-2, ..., 200-N) and a database (300) may be connected to a wired or wireless network to communicate. A wired or wireless network refers to a connection structure in which information can be exchanged between each node, such as a terminal or a server. Examples of such networks include, but are not limited to, the Internet, a LAN (local area network), a wireless LAN (local area network), a WAN (wide area network), a PAN (personal area network), 3G, 4G, 5G, Wi-Fi, Bluetooth, NFC, RFID, etc. Here, the plurality of user terminals (200-1, 200-2, ..., 200-N) are electronic devices capable of data communication through a wired or wireless network and may include various devices, such as a smartphone, a mobile phone, a tablet PC, a desktop, a laptop, etc., but are not limited thereto.

The support business recommendation service provided by the system (10) according to one embodiment is provided to multiple users, and can recommend support businesses that match the tendencies of each of the multiple users. Here, the user may be a participant in the support business recommendation service, and may correspond to an individual or a business that wishes to receive a recommendation for a support business that matches his or her tendencies among numerous support businesses. When multiple users provide the support businesses that they prefer to the device (100), and one of the multiple users requests a recommendation for a support business, the device (100) may recommend another support business similar to the support business preferred by the multiple user to the multiple user. The device (100) may analyze multiple support businesses preferred by multiple users, including the multiple user, through a recommendation model based on a collaborative filtering algorithm, and may recommend at least one support business among the multiple support businesses to the multiple user.

For example, the first to Nth users can use the support business recommendation service according to one embodiment by operating the first to Nth user terminals (200-1, 200-2, ..., 200-N), respectively. The second user can request the device (100) to recommend a support business suitable for the second user through the second user terminal (200-2). Previously, the first to Nth users, including the second user, can provide the device (100) with preferred support business information for their preferred support businesses through the first to Nth user terminals (200-1, 200-2, ..., 200-N). Alternatively, the preferred support business information of the first to Nth users can be stored in the database (300), and the device (100) can make a request to the database (300) to receive the corresponding preferred support business information.

Then, the device (100) can analyze the preferred support business information of the first to Nth users. In response to the recommendation request of the second user, the device (100) can provide the second user with information on the support business to be provided to the second user - recommended support business information. The process of deriving and providing the recommended support business information by the device (100) will be described in detail later.

FIG. 2 is a flowchart illustrating an operation of a device recommending a support project according to one embodiment.

Referring to FIG. 2, the operation of a device for recommending a support project according to one embodiment can be illustrated.

The device (100) can obtain support project information from multiple user terminals or databases (step S201). The support project information is information that describes and specifies the support project, and may include a unique number of the support project, a name of the support project, a hashtag, and/or a link. The unique number means a number assigned by the institution providing the support project for management purposes, the name of the support project means a name assigned by the institution to the support project, and a hashtag is a keyword or core word that describes the support project, and may have a form with a '#' next to it. The link may mean the address or source of the webpage where the support project is disclosed.

In addition, the device (100) can extract hashtags from the support project information (step S203). And the device (100) can calculate the similarity between the support projects from the hashtags (step S205). Since a hashtag is a kind of keyword and is a single or multiple keywords that describe the support projects, the similarity between the hashtags can be regarded as the similarity between the support projects. If two support projects are similar, they can include similar hashtags, and if the hashtags are similar to each other, the support projects can also be similar to each other.

In order to calculate the similarity between hashtags, the device (100) can use a word-to-vector algorithm. The device (100) can generate keywords included in a hashtag as vectors or numbers through the word-to-vector algorithm. That is, the device (100) can embed a hashtag. When a hashtag is expressed as a vector or number, similar hashtags can have similar vectors or numbers. If the vectors or numbers of two hashtags describing two support projects are similar to each other, the two hashtags can be considered similar to each other. That is, the similarity of the two hashtags can be high.

In addition, the device (100) can cluster support projects (step S207). Specifically, the device (100) can cluster support projects based on the similarity between hashtags and generate cluster information as a result thereof. For example, the device (100) can cluster support project information for multiple support projects in order of high similarity between hashtags. One or more support projects with high similarity between hashtags (high similarity between support projects) can be grouped into clusters. Preferably, the device (100) can cluster support projects into 3 to 4 clusters.

In addition, the device (100) can obtain user information for multiple users - for example, users 1 to N - and preferred support business information for support businesses preferred by multiple users (step S209). Here, the user information includes names or IDs that can identify multiple users, and the preferred support business information can include information on support businesses preferred by each user - for example, a unique number or name.

And the device (100) includes a recommendation model based on a collaborative filtering algorithm, and can derive recommended support business information to be recommended to one user among the plurality of users from user information and preferred support business information through the recommendation model (step S211). To this end, the recommendation model based on collaborative filtering can select other users similar to the one user from the user information and preferred support business information. This recommendation model can recommend at least one support business that is not preferred by the one user among the support businesses preferred by other users to the one user.

For example, if the first user prefers support projects A, support projects B, and support projects C, and the second user prefers support projects C, support projects D, and support projects E, the second user can request a recommendation of support projects from the device. Then, the recommendation model can select the first user who has a high similarity to the second user, and recommend support projects A and support projects B, which the second user did not prefer among the support projects preferred by the first user, to the second user. The recommendation model can derive information about support projects A and support projects B as recommended support project information. Extending this example, the recommendation model can select a user who has a high similarity to the second user from among the first to N users.

And the device (100) can adjust the derived recommended support business information by reflecting at least one of the support business information, the hashtag, and the similarity and cluster information between the support businesses derived from the hashtag (step S213). Specifically, the device can derive recommended support business information including multiple support businesses for one user, and can exclude or change some of the support businesses included in the derived recommended support business information based on at least one of the support business information, the hashtag, the similarity, and the cluster information. In the above example, the device can derive support businesses A and B to recommend to the second user. Support businesses A and B are similar to each other, but the similarity of support business A to support business C may be higher than that of support business B. The device can exclude support business B in the order of high similarity and include only support business A in the recommended support business information. Or, the device can include only support business B included in the same cluster as support business C, support business D, and support business E preferred by the second user in the recommended support business information.

Finally, the device can complete the recommendation task by providing the adjusted recommendation support business information to the user (step S215).

According to one embodiment, a device obtains preferred support business information from a plurality of users to be provided to a collaborative filtering-based recommendation model, wherein the preferred support business information may include support businesses that each user has designated as preferred within a certain period of time. The certain period of time may be a period from the present to a specific past time. For example, the device may obtain information on support businesses that the plurality of users have preferred within the past year from the time at which they provide the preferred support business information. Then, when there is a request from one of the plurality of users, the device may derive information on support businesses to be recommended to the user through the recommendation model - recommended support business information.

In addition, the preferred support business information provided by multiple users may include multiple support businesses preferred by each user. A user requesting a recommendation of a support business may also provide multiple support businesses preferred by the user to the device's recommendation model. Then, when the recommendation model derives recommended support business information for a user, the recommendation model may derive recommended support business information for each of the support businesses that the user has designated as preferred.

For example, users 1 to N can provide information on 10 support projects they have preferred over the past year - preferred support project information - to a recommendation model based on collaborative filtering. When a second user requests a recommendation of a support project from the device, the recommendation model can derive support projects for each of the 10 support projects preferred by the second user. Since there are 10 preferred support projects and 10 recommended support projects are derived for each, a total of 100 support projects can be derived for the second user. If all support projects preferred by users 1 to N are provided to the recommendation model and the recommendation model recommends support projects to the second user for all support projects, the computation may be large and it may take a lot of time. Therefore, the recommendation model performs collaborative filtering only on a large number of support projects preferred by multiple users during a certain period of time included in a certain period.

FIG. 3 is a diagram for explaining a method for recommending a support project according to one embodiment.

Referring to FIG. 3, in a method for recommending a support project according to one embodiment, a device may obtain support project information, extract a hashtag from the support project information, calculate a similarity between support projects, and perform clustering of the support projects. In addition, the device may obtain user information and preferred support project information of multiple users. The device may perform collaborative filtering on the user information and preferred support project information through a recommendation model. When one of the multiple users requests the device to recommend a support project, the recommendation model may derive recommended support project information to be recommended to the one user. In this derivation process, the recommendation model may adjust the derived recommended support project information by reflecting at least one of the support project information, the hashtag, the similarity, and the clustering information.

Meanwhile, collaborative filtering-based recommendation models can be tested to verify their performance. K-fold cross-validation can be used to test whether the recommendation model produces consistent performance results. In addition, quantitative analysis such as the time taken to make recommendations and/or MSE and statistical analysis such as t-test can be performed on the recommendation model.

Figure 4 is a configuration diagram of a device that recommends a support project according to one embodiment.

Referring to FIG. 4, a device (100) according to one embodiment may include a processor (110), a memory (120), a communication interface unit (130), a user interface unit (140), and a recommendation model (150). However, not all of the illustrated components are essential components. The device (100) may be implemented with more components than the illustrated components, and the device (100) may also be implemented with fewer components. The above components will be described below.

The communication interface unit (130) can perform communication with an external device. Specifically, the communication interface unit (130) can be connected to a network by wire or wirelessly to perform communication with an external device. Here, the external device can be a server and a terminal (such as a smartphone, tablet, or PC). The communication interface unit (130) can include a communication module that supports one of various wired and wireless communication methods. For example, the communication module can be in the form of a chipset, or a sticker/barcode including information required for communication. In addition, the communication module can be a short-range communication module or a wired communication module.

For example, the communication interface unit (130) may support at least one of wireless LAN, Wi-Fi (Wireless Fidelity), WFD (Wi-Fi Direct), Bluetooth, BLE (Bluetooth Low Energy), Wired Lan, NFC (Near Field Communication), Zigbee (Zigbee), infrared (IrDA, infrared Data Association), 3G, 4G, and 5G.

The user interface unit (140) may refer to a device that receives data from a user to control the device (100). The device (100) may control the user interface unit (140) to generate and output a user interface screen for receiving a predetermined command or data from a user. For example, the user interface unit (140) may include devices capable of receiving user input, such as a touch screen.

Software or a program may be stored in the memory (120). A command recommending a support project may be stored in the memory (120). In addition, at least one command or at least one program for operating the device (100) may be stored in the memory (120).

The processor (110) can control the overall operation of the device (100). The processor (110) may include at least one specialized processor corresponding to each function, or may be an integrated processor.

The processor (110) can execute a program stored in the memory (120), read data or a file stored in the memory (120), or store new data or a new file in the memory (120). The processor (110) can execute instructions stored in the memory (120).

By executing the instruction, the processor (130)

A method for obtaining support project information, extracting hashtags from the support project information, calculating a similarity between the support projects from the hashtags, clustering the support projects based on the similarity to create cluster information, obtaining user information on a plurality of users and preferred support project information on support projects preferred by the plurality of users, and deriving recommended support project information for recommendation to one of the plurality of users from the user information and the preferred support project information through a recommendation model using a collaborative filtering-based algorithm, adjusting the derived recommended support project information by reflecting at least one of the support project information, the hashtag, the similarity, and the cluster information, and providing the adjusted recommended support project information to the one user.

Here, the recommendation model (150) can obtain user information on multiple users and preferred support business information on support businesses preferred by the multiple users, and perform collaborative filtering on the user information and the preferred support business information. When one of the multiple users requests a recommendation for a support business, the recommendation model (150) can derive recommended support business information to recommend to the user.

The components or “units” or blocks or modules used in this embodiment may be implemented as software such as tasks, classes, subroutines, processes, objects, execution threads, programs performed in a certain area of memory, or as hardware such as FPGAs (field programmable gate arrays) or ASICs (application-specific integrated circuits), and may also be formed by a combination of the software and hardware. The components or “units” may be included in a computer-readable storage medium, and some of them may be distributed and distributed to a plurality of computers. In addition, one or more of the components may be implemented using one or more computing devices or parts thereof. Such devices may include, for example, personal computers, server computers, handheld or laptop devices, multiprocessor systems, microcontroller-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, cellular telephones, personal digital assistants (PDAs), gaming devices, printers, devices including set-top boxes, media centers, or other devices, automotive embedded or attached computing devices, other mobile devices, distributed computing environments including any of the above systems or devices, and the like.

Meanwhile, the disclosed embodiments may be implemented in the form of a recording medium storing a program and/or instructions executable by a computer. The instructions may be stored in the form of program codes, and when executed by a processor, may generate a program module to perform the operations of the disclosed embodiments. The recording medium may be implemented as a computer-readable recording medium. The computer-readable recording medium includes all kinds of recording media storing instructions that can be deciphered by a computer. For example, there may be a ROM (read only memory), a RAM (random access memory), a magnetic tape, a magnetic disk, a flash memory, an optical data storage device, etc.

The terms "include," "comprise," or "have" as used herein, unless otherwise specifically stated, imply that the corresponding component may be included, and therefore should be construed to include other components rather than to exclude other components. All terms, including technical or scientific terms, unless otherwise defined, have the same meaning as commonly understood by a person of ordinary skill in the art to which the present invention pertains. Commonly used terms, such as terms defined in a dictionary, should be interpreted as being consistent with their meaning in the context of the relevant art, and shall not be interpreted in an ideal or overly formal sense, unless expressly defined in the present invention.

The scope of protection of the present invention is not limited to the description and expression of the embodiments explicitly described above. In addition, it is added once again that the scope of protection of the present invention may not be limited due to obvious changes or substitutions in the technical field to which the present invention belongs.

Claims (6)

In the method by which the device recommends a support project,
Step for obtaining support project information;
A step of extracting hashtags from the above support project information;
A step of calculating the similarity between the above-mentioned support projects from hashtags;
A step of generating cluster information by clustering the support projects based on the above similarity;
A step of obtaining user information for multiple users and preferred support business information for preferred support businesses of the multiple users;
A step of deriving recommended support business information for recommendation to one of the plurality of users from the user information and the preferred support business information using a recommendation model based on a collaborative filtering-based algorithm;
A step of adjusting the derived recommended support project information by reflecting at least one of the above support project information, the hashtag, the similarity, and the cluster information; and
Including a step of providing the above adjusted recommendation support business information to the above user.
How to recommend a support project. In the first paragraph,
The steps for calculating the above similarity are:
Convert the above hashtags into vectors, and calculate the similarity between the above hashtags based on the converted vectors.
How to recommend a support project. In the first paragraph,
The above preferred support project information includes information on a number of support projects preferred by multiple users over a certain period of time.
How to recommend a support project. In the third paragraph,
The step of deriving the above recommended support business information through the above recommended model is:
For each of the above-mentioned support projects preferred by the above-mentioned users during the above-mentioned period, information on recommended support projects is derived.
How to recommend a support project. In the first paragraph,
The step of adjusting the above derived recommendation support business information is:
Excluding some of the support projects included in the recommended support project information derived based on at least one of the above support project information, the above hashtag, the above similarity, and the above cluster information.
How to recommend a support project. processor; and
comprising a memory storing instructions executable by the processor;
The above processor, by executing the above instructions,
Obtain information on support projects,
Extract hashtags from the above support project information,
Calculate the similarity between the above support projects from the hashtags,
Based on the above similarity, the above support projects are clustered to create cluster information.
Obtain user information for multiple users and preferred support business information for preferred support businesses of the multiple users,
A recommendation model using a collaborative filtering-based algorithm is used to derive recommended support business information for recommendation to one of the multiple users from the user information and the preferred support business information.
Adjust the derived recommended support project information by reflecting at least one of the above support project information, the above hashtag, the above similarity, and the above cluster information.
A device that recommends a support project by providing the above-mentioned adjusted recommendation support project information to the above-mentioned user.

Images