KR102612538B1 - Search enhancement method and device based on artificial neural network - Google Patents

Abstract

The present invention relates to an artificial neural network-based search enhancement method. The artificial neural network-based search enhancement method includes receiving a search request including a search keyword, determining at least one search method for performing a search among a plurality of search methods based on the received search keyword, and the determined at least and providing search results corresponding to the search request using one search method.

Description

Search enhancement method and device based on artificial neural network {SEARCH ENHANCEMENT METHOD AND DEVICE BASED ON ARTIFICIAL NEURAL NETWORK}

The present invention relates to an artificial neural network-based search enhancement method and device, and more specifically, to an artificial neural network-based search enhancement method and device for providing optimal search results by fusing various search methods.

Recently, with the development of smartphones, numerous searches are being performed through search engines. When a user enters a desired search word in the search box, the search engine extracts various keywords related to the entered search word through a search word dictionary and displays data about the extracted keywords on the screen. However, this keyword-based search method has the problem of providing accurate search results for keywords that are not predefined.

Meanwhile, a search method that performs a search by converting keywords into vectors is being studied. This vector-based search method has the advantage of providing similar search results even for keywords that are not predefined. However, because this vector-based search method uses a learned artificial neural network, it is difficult for users to directly edit or change the search results.

The present invention provides an artificial neural network-based search enhancement method to solve the above problems, a computer program stored in a computer-readable medium, a computer-readable medium storing the computer program, and a device (system).

The present invention may be implemented in various ways, including a method, an apparatus (system), a computer program stored in a computer-readable medium, or a computer-readable medium on which the computer program is stored.

According to an embodiment of the present invention, an artificial neural network-based search enhancement method performed by at least one processor includes receiving a search request including a search keyword, and selecting one of a plurality of search methods based on the received search keyword. It includes determining at least one search method for performing a search and providing search results corresponding to the search request using the determined at least one search method.

According to an embodiment of the present invention, the step of determining at least one search method among a plurality of search methods includes at least one of a first search method using a vector based on a received search keyword and a second search method using a keyword. It includes the step of determining a search method.

According to an embodiment of the present invention, the step of determining at least one search method among a first search method using a vector and a second search method using a keyword includes a first reliability score of the first search method for the search keyword, and It includes calculating a second reliability score of a second search method and determining at least one search method based on whether the calculated first reliability score or the second reliability score is greater than or equal to a first threshold.

According to an embodiment of the present invention, the step of calculating the first reliability score of the first search method and the second reliability score of the second search method includes providing a search keyword to the learned artificial neural network model to determine the first reliability score of the first search method. It includes calculating a first reliability score and a second reliability score of the second search method.

According to one embodiment of the present invention, when at least a portion of the first reliability score and the second reliability score is less than or equal to the second threshold, updating a search term dictionary associated with a second search method using a keyword based on the search keyword. Includes more.

According to an embodiment of the present invention, the first search method includes a search method that extracts similar vector expressions by comparing vectors corresponding to search keywords with data stored in a vector database.

According to an embodiment of the present invention, the second search method includes a search method using an artificial neural network-based search term dictionary that is automatically updated using content collected through a crawling algorithm.

According to one embodiment of the present invention, the artificial neural network-based search term dictionary includes a thesaurus, a stop word dictionary, and a user dictionary.

A computer program stored in a computer-readable recording medium is provided to execute the method according to an embodiment of the present invention on a computer.

A computing device according to an embodiment of the present invention includes a communication module, a memory, and at least one processor connected to the memory and configured to execute at least one computer-readable program included in the memory. At least one program receives a search request including a search keyword, determines at least one search method among a plurality of search methods based on the received search keyword, and responds to the search request using the determined at least one search method. Contains commands to provide corresponding search results.

In various embodiments of the present invention, a computing device can significantly improve search accuracy by determining an optimal search method according to a search keyword and providing search results searched using the determined search method to the user.

In various embodiments of the present invention, the computing device can derive optimal search results by appropriately utilizing vector-based search methods and keyword-based search methods, and when search reliability is low, the corresponding search keywords are appropriately updated in the search term dictionary. Therefore, even if the same search keyword is received in the future, precise search results can be provided through a keyword-based search method.

In various embodiments of the present invention, the computing device can automatically extract and classify the keyword and update the search term dictionary even when a new keyword is used on the web. Additionally, if the search keyword entered by the user is a new keyword that is not defined in the dictionary, the computing device can automatically extract and classify the keyword to update the search term dictionary and improve search performance.

In various embodiments of the present invention, when a new keyword is created, keyword classification can be simply performed using an arbitrary machine learning model and/or algorithm without a person directly performing keyword classification, and the search term dictionary can be updated based on this. there is.

The effects of the present invention are not limited to the effects mentioned above, and other effects not mentioned are clear to a person skilled in the art (referred to as a “person skilled in the art”) in the technical field to which the present invention pertains from the description of the claims. It will be understandable.

Embodiments of the present invention will be described with reference to the accompanying drawings described below, in which like reference numerals represent like elements, but are not limited thereto.
1 is a diagram illustrating an example of a computing device that performs search augmentation through various search methods according to an embodiment of the present invention.
Figure 2 is a functional block diagram showing the internal configuration of a computing device according to an embodiment of the present invention.
Figure 3 is a diagram illustrating an example of a computing device classifying keywords according to an embodiment of the present invention.
Figure 4 is a diagram illustrating an example in which a computing device calculates a relevance score of a first keyword and another keyword when updating a search term dictionary according to an embodiment of the present invention.
Figure 5 is an exemplary diagram showing an artificial neural network model according to an embodiment of the present invention.
Figure 6 is a diagram showing an example of an artificial neural network-based search enhancement method according to an embodiment of the present invention.
Figure 7 is a block diagram showing the internal configuration of a computing device according to an embodiment of the present invention.

Hereinafter, specific details for implementing the present invention will be described in detail with reference to the attached drawings. However, in the following description, detailed descriptions of well-known functions or configurations will be omitted if there is a risk of unnecessarily obscuring the gist of the present invention.

In the accompanying drawings, identical or corresponding components are given the same reference numerals. Additionally, in the description of the following embodiments, overlapping descriptions of identical or corresponding components may be omitted. However, even if descriptions of components are omitted, it is not intended that such components are not included in any embodiment.

Advantages and features of the disclosed embodiments and methods for achieving them will become clear by referring to the embodiments described below in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below and may be implemented in various different forms, but the present embodiments only serve to ensure that the present invention is complete and that the scope of the invention is conveyed to those skilled in the art. It is provided only for complete information.

Terms used in this specification will be briefly described, and the disclosed embodiments will be described in detail. The terms used in this specification are general terms that are currently widely used as much as possible while considering the function in the present invention, but this may vary depending on the intention or precedent of a technician working in the related field, the emergence of new technology, etc. In addition, in certain cases, there are terms arbitrarily selected by the applicant, and in this case, the meaning will be described in detail in the description of the relevant invention. Therefore, the terms used in the present invention should be defined based on the meaning of the term and the overall content of the present invention, rather than simply the name of the term.

In this specification, singular expressions include plural expressions, unless the context clearly specifies the singular. Additionally, plural expressions include singular expressions, unless the context clearly specifies plural expressions. When it is said that a certain part includes a certain element throughout the specification, this does not mean excluding other elements, but may further include other elements, unless specifically stated to the contrary.

In the present invention, terms such as "comprise", "comprising", etc. may indicate the presence of features, steps, operations, elements and/or components, but may indicate that such terms include one or more other functions, It does not preclude the addition of steps, operations, elements, components and/or combinations thereof.

In the present invention, when a specific component is referred to as being “coupled,” “combined,” “connected,” or “reacting” with any other component, the specific component is directly bonded, combined, and/or connected to the other component. Alternatively, it may be connected or react, but is not limited thereto. For example, one or more intermediate components may exist between a particular component and another component. Additionally, in the present invention, “and/or” may include each of one or more listed items or a combination of at least a portion of one or more items.

In the present invention, terms such as “first” and “second” are used to distinguish specific components from other components, and the components described above are not limited by these terms. For example, the “first” component may be an element of the same or similar form as the “second” component.

In the present invention, a “search word dictionary” is a database in which a plurality of keywords are stored and may include a thesaurus, a stop word dictionary, a user dictionary, etc. For example, a thesaurus is used to manage synonyms of specific keywords and may include a plurality of groups of similar search terms. Additionally, each similar search term group may include multiple keywords. According to one embodiment, when a similar search word group includes the keywords "A, B, C, D" and A is entered as a search word, the keywords in the similar search word group including A are "A, B, C , D" search results may be output. Additionally, a stopword dictionary may be used to manage keywords related to meaningless parts of speech such as particles and prepositions, profanity, slang, and derogatory words. Additionally, the user dictionary may include keywords related to compound words such as new words, or may be used to manage the priority of keywords.

In the present invention, “search enhancement” may refer to a search technique that derives optimal search results using at least some of a plurality of search methods.

In the present invention, the “search method using vectors” involves embedding the input search keyword into a vector expression and then comparing the similarity using the distance between the embedded vector expression and the embedding space of dense vectors stored in the vector database. It can indicate the search method you are looking for.

In the present invention, the "search method using keywords" is a search method that uses a search word dictionary with predefined keywords, and is a search method that uses a search word dictionary managed by the user and a search word dictionary that is automatically updated based on an artificial neural network model. It may include an artificial neural network-based search method used.

In the present invention, the “reliability score” is a value indicating the search accuracy of a specific search method corresponding to a search keyword, and can be calculated, for example, by an artificial neural network model using various parameters.

FIG. 1 is a diagram illustrating an example of a computing device 100 that performs search augmentation through various search methods according to an embodiment of the present invention. According to one embodiment, the computing device 100 may refer to any device for outputting and providing search results by selecting an optimal search method based on the search keyword 102 input by any user. That is, unlike a typical system that provides search results through a single search engine, the computing device 100 can provide optimal search results using at least some of various search methods.

According to one embodiment, the computing device 100 may receive a search request including the search keyword 102 from any user (or user terminal). In this case, the computing device 100 may analyze the received search keyword 102 and determine at least one search method to provide precise search results corresponding to the search keyword 102. For example, the first search method 104 may be a search method using a vector, and the second search method 106 may be a traditional search method using a predetermined search word dictionary based on keywords. , The third search method 108 may be a keyword-based search method that uses a search term dictionary that is automatically updated based on an artificial neural network. That is, when receiving a search request, the computing device 100 corresponds at least one of the first search method 104, the second search method 106, and the third search method 108 to the corresponding search keyword 102. You can decide by searching method.

According to one embodiment, the first search method 104 using vectors refers to a search method that provides search results corresponding to the search keyword 102 using an interactive artificial intelligence service (e.g., Chat GPT service), etc. can do. In the case of this first search method 104, there is an advantage in that search results can be provided by analyzing the input search keyword even when a new search keyword not defined in a dictionary, such as a search term dictionary, is input. However, for a specific search keyword, the search results of the first search method 104 may be inaccurate, and it may be difficult for the user to modify the search results. In addition, the second search method 106 and/or the third search method 108 based on keywords may provide matching search results using a search word dictionary, but provide accurate search results for search keywords not included in the search word dictionary. Doesn't provide results.

As such, the search results and accuracy are different depending on the search method, so the computing device 100 uses the first search method 104, the second search method 106, and the third search method 108 according to the search keyword 102. By selecting and using at least some of the above, search results corresponding to the search request can be provided to the user. With this configuration, the computing device 100 can significantly improve search accuracy by determining the optimal search method according to the search keyword 102 and providing search results searched using the determined search method to the user.

Figure 2 is a functional block diagram showing the internal configuration of the computing device 100 according to an embodiment of the present invention. As described above, the computing device 100 is an arbitrary device for selecting an optimal search method based on the search keyword 102 and outputting and providing search results, including a communication unit 210 and a search method determination unit 220. ), a dictionary update unit 230, etc. Additionally, the computing device 100 may communicate with the vector database 240, the search term dictionary 250, etc., and exchange data and/or information required for search.

According to one embodiment, the communication unit 210 communicates with a user terminal, etc. and may receive a search request including a search keyword or provide search results. For example, when receiving a search request from a user terminal, the communication unit 210 may provide the search keyword included in the search request to the search method determination unit 220. Then, the communication unit 210 may provide search results searched using a random search method to the corresponding user terminal.

According to one embodiment, the search method determination unit 220 may determine at least one search method for performing a search among a plurality of search methods based on the received search keyword. For example, the search method determination unit 220 may determine at least one search method among a first search method using a vector and a second search method using a keyword based on the received search keyword.

In this case, the search method determination unit 220 calculates a first reliability score of the first search method and a second reliability score of the second search method for the search keyword, and the calculated first reliability score or the second reliability score is At least one search method may be determined based on whether the value is greater than or equal to the first threshold. For example, when both the first reliability score and the second reliability score are equal to or greater than the first threshold, the search method determination unit 220 may use both the first search method and the second search method. In another example, when the first reliability score is greater than or equal to the first threshold and the second reliability score is less than the first threshold, the search method determination unit 220 may use only the first search method.

According to one embodiment, the search method determination unit 220 provides a search keyword to a learned artificial neural network model (e.g., a first artificial neural network model) to provide a first reliability score of the first search method and a first reliability score of the second search method. 2 Reliability scores can be calculated. For example, when a search keyword is obtained, computing device 100 may convert the search keyword into a vector representation. Then, the artificial neural network model can calculate the similarity between the vector expression corresponding to the search keyword and the vector expressions stored on the vector database 240. In this case, the artificial neural network model can calculate the first reliability score of the first search method higher as the similarity between vector expressions is calculated higher. In another example, when a search keyword is obtained, the artificial neural network model may calculate the similarity between the search keyword and keywords included in the search word dictionary 250. In this case, the artificial neural network model can calculate the second reliability score of the second search method higher as the similarity between keywords is calculated higher.

According to one embodiment, the dictionary update unit 230 may update the search term dictionary 250 associated with the second search method. For example, the dictionary update unit 230 may obtain keywords from content included on the web and update the search term dictionary 250 based on the obtained keywords. For example, the computing device 100 may collect content included on the web using a crawling algorithm and extract keywords from the collected content. In this case, the computing device 100 compares the keyword with a plurality of keywords included in the search word dictionary 250, and then extracts the keyword from the collected content when it is determined that the keyword is a new keyword that is not in the search word dictionary 250. and the search term dictionary 250 can be updated.

Additionally or alternatively, when the first reliability score and the second reliability score are less than or equal to the second threshold, the dictionary update unit 230 updates the search term dictionary 250 associated with a second search method using a keyword based on the search keyword. It can be updated. That is, if it is determined that it is difficult to provide accurate search results even with the vector-based first search method and/or the keyword-based second search method, the dictionary update unit 230 enters the corresponding search keyword into the search word dictionary 250. You can use it by adding it.

In FIG. 2 , each functional configuration included in the computing device 100 is separately described and described in detail, but this is only to aid understanding of the invention, and two or more functions may be performed in one computing device. With this configuration, the computing device 100 can derive optimal search results by appropriately utilizing the vector-based search method and the keyword-based search method, and when search reliability is low, the corresponding search keyword is entered into the search term dictionary ( 250), it is possible to provide precise search results through a keyword-based search method even if the same search keyword is received in the future.

FIG. 3 is a diagram illustrating an example of how the computing device 100 classifies keywords according to an embodiment of the present invention. As described above, the computing device 100 is updated using content collected through a crawling algorithm, or is updated using an artificial neural network-based search word dictionary that is automatically updated using search keywords with a reliability score below a threshold. You can perform a search. Here, the artificial neural network-based search term dictionary may include a thesaurus, a stop word dictionary, and a user dictionary.

According to one embodiment, the computing device 100 may obtain keywords 312 based on the second artificial neural network model 310 and perform keyword classification 314. For example, the computing device 100 may obtain keywords from content included on the web. Here, content is a post that exists online and may include SNS posts, Internet news and/or articles, blog posts, etc. That is, the computing device 100 can collect content included on the web using a crawling algorithm and extract keywords 312 from the collected content. For example, the computing device 100 compares the keyword 312 with a plurality of keywords included in the search term dictionary, and when it is determined that the keyword 312 is a new keyword that is not in the search term dictionary, the computing device 100 selects the keyword ( 312) can be extracted.

Additionally or alternatively, computing device 100 may extract search keywords entered by the user as keywords 312 . For example, when the first reliability score of the first search method and the second reliability score of the second search method for the search keyword are less than or equal to the second threshold, the computing device 100 may use the corresponding search keyword to update the search word dictionary. It can be extracted by keyword (312).

According to one embodiment, the computing device 100 divides the extracted keyword into morpheme units and provides the morpheme divided and extracted from the keyword to the learned second artificial neural network model 310 to perform keyword classification 314. can do. Here, the second artificial neural network model 310 is a classifier or classification model learned to classify new unlabeled information based on relevance to labeled data and/or It can refer to a language model (e.g. BERT model, etc.) for analyzing the meaning of keywords in context. For example, when receiving a new keyword 312, the second artificial neural network model 310 may classify the keyword 312 into a group of similar search terms of other keywords.

According to one embodiment, when the second artificial neural network model 310 determines that the morpheme associated with the keyword 312 is a typo of another keyword, the second artificial neural network model 310 may classify the keyword 312 into a similar search word group of another keyword. Additionally or alternatively, if the second artificial neural network model 310 determines that the morpheme associated with the keyword 312 is a synonym of another keyword, the second artificial neural network model 310 may classify the keyword 312 into a group of similar search terms of the other keyword. Additionally or alternatively, the second artificial neural network model 310 calculates a relationship score of the morpheme associated with the keyword 312 and other keywords, and if the calculated relationship score is greater than or equal to a predetermined standard, the keyword 312 ) can be classified into a group of similar search terms of other keywords. Here, the relevance score may be a probability value indicating the relevance of the keyword 312 and other keywords.

Additionally or alternatively, when receiving a new keyword 312, the second artificial neural network model 310 may classify the keyword 312 as a stopword. For example, the second artificial neural network model 310 may calculate the stop word score of the keyword 312 based on the context in the sentence in which the keyword 312 is used, and if the stop word score is higher than a predetermined standard, the second artificial neural network model 310 may calculate the stop word score of the keyword 312. Keywords 312 can be classified as stop words. Here, the stop word score may be a probability value indicating a negative connotation, and may be calculated high when, for example, it is used in the same sentence as another slang word.

Additionally or alternatively, when receiving a new keyword 312, the second artificial neural network model 310 may classify the keyword 312 as a compound word such as a new word. For example, the second artificial neural network model 310 can calculate a new word score based on the form in which the keyword 312 is used in the content, and if the new word score is higher than a predetermined standard, the keyword 312 is used as a new word. It can be classified as: Here, the new word score may represent a probability value that the keyword 312 is a new compound word. For example, the more the keyword 312 is repeatedly used in the form of a compound word within the content, the higher it can be calculated.

When keywords are classified in this way, the computing device 100 may create or update a search word dictionary to include the classification results. With this configuration, the computing device 100 can automatically extract and classify the keyword and update the search word dictionary even when a new keyword is used on the web. Additionally, if the search keyword entered by the user is a new keyword that is not defined in the dictionary, the computing device 100 may automatically extract and classify the keyword to update the search term dictionary and improve search performance.

FIG. 4 is a diagram illustrating an example in which the computing device 100 calculates a relevance score of a keyword different from the first keyword 410 when updating a search term dictionary according to an embodiment of the present invention. As described above, the computing device 100 calculates the relevance score of the morpheme associated with the first keyword 410 and the other keyword, and when the calculated relevance score is greater than or equal to a predetermined standard, the computing device 100 uses the first keyword 410 as another keyword. It can be classified into groups of similar search words. For example, the computing device 100 may use a random machine learning model and/or The relevance score can be calculated using an algorithm.

As shown, the computing device 100 receives the first keyword 410 as an input and uses a plurality of other keywords, that is, the second keyword 420, the third keyword 430, the fourth keyword 440, etc. Relevance scores (422, 432, 442) can be output. For example, the probability value that the first keyword 410 is associated with the second keyword 420 may be 0.7, the probability value that the first keyword 410 is associated with the third keyword 430 may be 0.2, The probability value that the first keyword 410 is associated with the fourth keyword 440 may be 0.1. That is, the computing device 100 may determine that the first keyword 410 is highly related to the second keyword 420.

With this configuration, when a new keyword is created, keyword classification can be simply performed using an arbitrary machine learning model and/or algorithm without a person directly performing keyword classification, and the search term dictionary can be updated based on this. .

Figure 5 is an exemplary diagram showing an artificial neural network model 500 according to an embodiment of the present invention. The artificial neural network model 500 is an example of a machine learning model, and in machine learning technology and cognitive science, it is a statistical learning algorithm implemented based on the structure of a biological neural network or a structure for executing the algorithm.

According to one embodiment, the artificial neural network model 500 has nodes, which are artificial neurons that form a network through the combination of synapses, as in a biological neural network, repeatedly adjusting the weights of synapses to determine the correct input corresponding to a specific input. By learning to reduce the error between the output and the inferred output, a machine learning model with problem-solving capabilities can be represented. For example, the artificial neural network model 500 may include a random probability model, a neural network model, etc. used in artificial intelligence learning methods such as machine learning and deep learning.

The artificial neural network model 500 is implemented as a multilayer perceptron (MLP) consisting of multiple layers of nodes and connections between them. The artificial neural network model 500 according to this embodiment can be implemented using one of various artificial neural network model structures including MLP. As shown in FIG. 5, the artificial neural network model 500 includes an input layer 520 that receives an input signal or data 510 from the outside, and an output layer that outputs an output signal or data 550 corresponding to the input data. (540), located between the input layer 520 and the output layer 540, receives n signals from the input layer 520, extracts the characteristics, and transmits them to the output layer 540 (where n is a positive integer) It consists of hidden layers (530_1 to 530_n). Here, the output layer 540 receives signals from the hidden layers 530_1 to 530_n and outputs them to the outside.

The learning method of the artificial neural network model 500 includes a supervised learning method that learns to optimize problem solving by inputting a teacher signal (correct answer), and an unsupervised learning method that does not require a teacher signal. ) There is a way. According to one embodiment, input variables of the artificial neural network model 500 may include search keywords, etc. When the above-described input variable is input through the input layer 520, the output variable output from the output layer 540 of the artificial neural network model 500 is at least one to provide the optimal search result corresponding to the search keyword. It can be a search method, etc.

In this way, a plurality of input variables and a plurality of output variables corresponding to the input layer 520 and the output layer 540 of the artificial neural network model 500 are matched, respectively, and the input layer 520, the hidden layers 530_1 to 530_n, and By adjusting the synapse values between nodes included in the output layer 540, learning can be performed so that the correct output corresponding to a specific input can be extracted. Through this learning process, the characteristics hidden in the input variables of the artificial neural network model 500 can be identified, and the nodes of the artificial neural network model 500 can be connected to reduce the error between the output variable calculated based on the input variable and the target output. You can adjust the synapse values (or weights) between them.

According to one embodiment, the artificial neural network model 500 may classify the search keyword by receiving morphemes divided and extracted from the search keyword. For example, the artificial neural network model 500 may classify search keywords into similar search word groups of other specific keywords stored in the search word dictionary. In another example, the artificial neural network model 500 may classify keywords as stop words or as compound words such as new words.

Figure 6 is a diagram showing an example of an artificial neural network-based search enhancement method 600 according to an embodiment of the present invention. The artificial neural network-based search augmentation method 600 may be performed by at least one processor (eg, at least one processor of a computing device). The artificial neural network-based search enhancement method 600 may be initiated by the processor receiving a search request including a search keyword (S610). For example, the processor may receive a search request from any user terminal.

The processor may determine at least one search method for performing a search among a plurality of search methods based on the received search keyword (S620). For example, the processor may determine at least one search method among a first search method using a vector and a second search method using a keyword based on the received search keyword.

According to one embodiment, the processor calculates a first reliability score of a first search method and a second reliability score of a second search method for a search keyword, and the calculated first reliability score or the second reliability score is set to a first threshold. At least one search method may be determined based on whether the value is greater than or equal to the value. For example, the processor may provide a search keyword to the learned artificial neural network model to calculate a first reliability score of the first search method and a second reliability score of the second search method.

Then, the processor may provide a search result corresponding to the search request using the determined at least one search method (S630). That is, the processor compares vectors corresponding to search keywords with data stored in a vector database to extract similar vector expressions, and an artificial neural network-based search term that is automatically updated using content collected through a crawling algorithm. Search results can be extracted and provided using at least some of the second search methods using a dictionary.

FIG. 7 is a block diagram showing the internal configuration of the computing device 100 according to an embodiment of the present invention. The computing device 100 may include a memory 710, a processor 720, a communication module 730, and an input/output interface 740. As shown in FIG. 7, the computing device 100 includes a communication module ( 730) may be configured to communicate information and/or data through a network.

Memory 710 may include any non-transitory computer-readable recording medium. According to one embodiment, the memory 710 is a non-permanent mass storage device such as random access memory (RAM), read only memory (ROM), disk drive, solid state drive (SSD), flash memory, etc. mass storage device). As another example, non-perishable mass storage devices such as ROM, SSD, flash memory, disk drive, etc. may be included in the computing device 100 as a separate persistent storage device that is distinct from memory. Additionally, an operating system and at least one program code may be stored in the memory 710.

These software components may be loaded from a computer-readable recording medium separate from the memory 710. Recording media readable by such a separate computer may include recording media directly connectable to the computing device 100, for example, floppy drives, disks, tapes, DVD/CD-ROM drives, memory cards, etc. It may include a computer-readable recording medium. As another example, software components may be loaded into the memory 710 through the communication module 730 rather than a computer-readable recording medium. For example, at least one program may be loaded into memory 710 based on a computer program installed by files provided through the communication module 730 by developers or a file distribution system that distributes installation files for applications. You can.

The processor 720 may be configured to process instructions of a computer program by performing basic arithmetic, logic, and input/output operations. Commands may be provided to another user terminal (not shown) or another external system by the memory 710 or the communication module 730.

The communication module 730 may provide a configuration or function for a user terminal (not shown) and the computing device 100 to communicate with each other through a network, and the computing device 100 may be configured to communicate with an external system (e.g., a separate cloud system). etc.) may provide a configuration or function for communication. For example, control signals, commands, data, etc. provided under the control of the processor 720 of the computing device 100 pass through the communication module 730 and the network to the user terminal and/or the communication module of the external system. and/or transmitted to an external system.

Additionally, the input/output interface 740 of the computing device 100 may be connected to the computing device 100 or may be a means for interfacing with a device (not shown) for input or output that the computing device 100 may include. . In FIG. 7 , the input/output interface 740 is shown as an element configured separately from the processor 720, but the present invention is not limited thereto, and the input/output interface 740 may be included in the processor 720. Computing device 100 may include more components than those of FIG. 7 . However, there is no need to clearly show most prior art components.

The processor 720 of the computing device 100 may be configured to manage, process, and/or store information and/or data received from a plurality of user terminals and/or a plurality of external systems.

The above-described method and/or various embodiments may be implemented with digital electronic circuitry, computer hardware, firmware, software, and/or combinations thereof. Various embodiments of the present invention are executed by a data processing device, for example, one or more programmable processors and/or one or more computing devices, or as a computer program stored in a computer-readable recording medium and/or a computer-readable recording medium. It can be implemented. The above-described computer program may be written in any form of programming language, including compiled language or interpreted language, and may be distributed in any form such as a stand-alone program, module, or subroutine. A computer program may be distributed via a single computing device, multiple computing devices connected through the same network, and/or multiple computing devices distributed so as to be connected through multiple different networks.

The above-described method and/or various embodiments include one or more processors configured to execute one or more computer programs that process, store, and/or manage any functions, functions, etc., by operating on input data or generating output data. It can be performed by . For example, the method and/or various embodiments of the present invention may be performed by special purpose logic circuits such as a Field Programmable Gate Array (FPGA) or Application Specific Integrated Circuit (ASIC), and the method and/or various embodiments of the present invention may An apparatus and/or system for performing embodiments may be implemented as a special purpose logic circuit, such as an FPGA or ASIC.

The one or more processors executing the computer program may include a general-purpose or special-purpose microprocessor and/or one or more processors in any type of digital computing device. The processor may receive instructions and/or data from each of read-only memory and random access memory, or may receive instructions and/or data from read-only memory and random access memory. In the present invention, components of a computing device performing methods and/or embodiments may include one or more processors for executing instructions and one or more memory devices for storing instructions and/or data.

According to one embodiment, a computing device may exchange data with one or more mass storage devices for storing data. For example, a computing device may receive data from and/or transmit data to a magnetic disc or optical disc. Computer-readable storage media suitable for storing instructions and/or data associated with a computer program include semiconductor memory devices such as EPROM (Erasable Programmable Read-Only Memory), EEPROM (Electrically Erasable PROM), and flash memory devices. It may include, but is not limited to, any form of non-volatile memory. For example, computer-readable storage media may include magnetic disks such as internal hard disks or removable disks, magneto-optical disks, CD-ROM, and DVD-ROM disks.

To provide interaction with a user, the computing device may include a display device (e.g., cathode ray tube (CRT), liquid crystal display (LCD), etc.) for presenting or displaying information to the user and a display device (e.g., cathode ray tube (CRT), liquid crystal display (LCD), etc.) for displaying or presenting information to the user. It may include, but is not limited to, a pointing device (e.g., keyboard, mouse, trackball, etc.) capable of providing input and/or commands. That is, the computing device may further include any other types of devices for providing interaction with the user. For example, a computing device may provide any form of sensory feedback to a user for interaction with the user, including visual feedback, auditory feedback, and/or tactile feedback. In response, the user can provide input to the computing device through various gestures such as sight, voice, and movement.

In the present invention, various embodiments may be implemented in a computing system that includes a back-end component (e.g., a data server), a middleware component (e.g., an application server), and/or a front-end component. In this case, the components may be interconnected by any form or medium of digital data communication, such as a communications network. For example, a communication network may include a Local Area Network (LAN), a Wide Area Network (WAN), etc.

Computing devices based on example embodiments described herein may be implemented using hardware and/or software configured to interact with a user, including a user device, user interface (UI) device, user terminal, or client device. You can. For example, a computing device may include a portable computing device, such as a laptop computer. Additionally or alternatively, computing devices include personal digital assistants (PDAs), tablet PCs, game consoles, wearable devices, internet of things (IoT) devices, virtual reality (VR) devices, AR (augmented reality) devices, etc. may be included, but are not limited thereto. Computing devices may further include other types of devices configured to interact with a user. Additionally, the computing device may include a portable communication device (eg, a mobile phone, smart phone, wireless cellular phone, etc.) suitable for wireless communication over a network, such as a mobile communication network. The computing device may wirelessly connect to a network server using wireless communication technologies and/or protocols, such as radio frequency (RF), microwave frequency (MWF), and/or infrared ray frequency (IRF). It can be configured to communicate with.

The various embodiments herein, including specific structural and functional details, are illustrative. Accordingly, embodiments of the present invention are not limited to those described above and may be implemented in various other forms. Additionally, the terms used in the present invention are intended to describe some embodiments and are not to be construed as limiting the embodiments. For example, the singular forms of words and the like may be construed to include the plural, unless the context clearly indicates otherwise.

In the present invention, unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by a person of ordinary skill in the technical field to which these concepts belong. . Additionally, commonly used terms, such as those defined in dictionaries, should be interpreted as having meanings consistent with their meaning in the context of the relevant technology.

Although the present invention has been described in relation to some embodiments in this specification, various modifications and changes may be made without departing from the scope of the present invention as understood by those skilled in the art. Additionally, such modifications and changes should be considered to fall within the scope of the claims appended hereto.

100: computing device
102: Search keyword
104: First search method
106: Second search method

Claims (17)

An artificial neural network-based search augmentation method performed by at least one processor, comprising:
Receiving a search request including a search keyword;
analyzing the received search keyword to determine at least one search method among a plurality of search methods for providing reliable search results for the search keyword; and
providing search results corresponding to the search request using the determined at least one search method;
Including,
The plurality of search methods include a first search method and a second search method,
The step of determining at least one search method to provide reliable search results for the search keyword among the plurality of search methods includes:
calculating a first reliability score of the first search method and a second reliability score of the second search method for the search keyword; and
determining the at least one search method based on whether the calculated first reliability score or the second reliability score is greater than or equal to a first threshold;
An artificial neural network-based search augmentation method including.
According to paragraph 1,
The step of determining at least one search method to provide reliable search results for the search keyword among the plurality of search methods includes:
determining at least one search method among a plurality of search methods, including a first search method using a vector and a second search method using a keyword, based on the received search keyword;
An artificial neural network-based search augmentation method including.
delete According to paragraph 1,
The step of calculating the first reliability score of the first search method and the second reliability score of the second search method,
providing the search keyword to a learned artificial neural network model to calculate a first reliability score of the first search method and a second reliability score of the second search method;
An artificial neural network-based search augmentation method including.
According to paragraph 1,
When at least a portion of the first reliability score and the second reliability score is less than or equal to a second threshold, updating a search word dictionary associated with a second search method using the keyword based on the search keyword;
An artificial neural network-based search augmentation method further comprising:
According to paragraph 2,
The first search method is,
An artificial neural network-based search enhancement method comprising a search method for extracting similar vector expressions by comparing vectors corresponding to the search keywords with data stored in a vector database.
According to paragraph 2,
The second search method is,
An artificial neural network-based search enhancement method, including a search method using an artificial neural network-based search term dictionary that is automatically updated using content collected through a crawling algorithm.
In clause 7,
The artificial neural network-based search word dictionary is,
Artificial neural network-based search augmentation method, including thesaurus, stopword dictionary, and user dictionary.
A computer program stored in a computer-readable recording medium for executing the method according to any one of claims 1, 2, and 4 to 8 on a computer.
As a computing device,
communication module;
Memory; and
At least one processor connected to the memory and configured to execute at least one computer-readable program included in the memory
Including,
The at least one program is,
Receive a search request containing search keywords,
Analyzing the received search keyword to determine at least one search method among a plurality of search methods for providing reliable search results for the search keyword,
Includes instructions for providing search results corresponding to the search request using the determined at least one search method,
The plurality of search methods include a first search method and a second search method,
Calculating a first reliability score of the first search method and a second reliability score of the second search method for the search keyword,
The computing device further includes instructions for determining the at least one search method based on whether the calculated first reliability score or the calculated second reliability score is greater than or equal to a first threshold.
According to clause 10,
The at least one program is,
The computing device further includes instructions for determining at least one search method among a plurality of search methods, including a first search method using a vector and a second search method using a keyword, based on the received search keyword.
delete According to clause 10,
The at least one program is,
A computing device further comprising instructions for calculating a first reliability score of the first search method and a second reliability score of the second search method by providing the search keyword to a learned artificial neural network model.
According to clause 10,
The at least one program is,
Computing further comprising instructions for updating a search word dictionary associated with a second search method using the keyword based on the search keyword when at least a portion of the first trust score and the second trust score is less than or equal to a second threshold. Device.
According to clause 11,
The first search method is,
Computing device comprising a search method for extracting similar vector expressions by comparing vectors corresponding to the search keywords with data stored in a vector database.
According to clause 11,
The second search method is,
A computing device including a search method using an artificial neural network-based search word dictionary that is automatically updated using content collected through a crawling algorithm.
According to clause 16,
The artificial neural network-based search word dictionary is,
A computing device, including a thesaurus, a stopword dictionary, and a user dictionary.

Images