Device, method, and recording medium for displaying review information of item

Abstract

This invention proposes a method for displaying review information of an item. One embodiment of the method may include the following steps: obtaining a user's connection request for a first page from a user terminal, the first page including detailed information of a target item; identifying one or more review items corresponding to the target item; extracting one or more review keywords based on the one or more review items; generating first page information for outputting the first page in such a way that at least a portion of the first page includes a review keyword area for displaying the one or more review keywords; and transmitting the first page information to the user terminal; wherein each of the one or more review items corresponding to the target item includes the target item's review score information and detailed evaluation information.

Description

[Technical Field]

[0001] This invention relates to a technique for displaying review information of an item. Specifically, it relates to a technique for displaying review information that can be provided to a user along with detailed information about the item. [Previous Technology]

[0002] With the widespread use of the Internet, e-commerce is constantly expanding. This phenomenon is accompanied by a decline in the rate of purchasing goods through traditional offline stores, and a sharp increase in the rate of purchasing goods through e-commerce platforms using computers or smartphones.

[0003] E-commerce services can provide product review information, that is, users' thoughts on the products. Users can obtain the information they want about the products more intuitively and quickly based on this review information. This can improve the user experience of using e-commerce services and ultimately increase the likelihood of being induced to purchase the products.

[0004] However, when there is an excessive amount of review information, or when the review information is complex, lengthy, or unclear, users may find it difficult to find the information they need. That is, users of e-commerce services may find it difficult to access review information intuitively and easily. This may result in users not obtaining sufficient information about the item, thus failing to establish the trust necessary to make a purchase. [Summary of the Invention]

[0005] [Problem to be solved by the invention] There is a need for a technology to improve the accessibility of comment information for users.

[0006] There is a need for a technology that can encourage users of e-commerce services to more easily access product reviews and information, while reducing user fatigue caused by information overload. [Technical means to solve the problem]

[0007] This invention proposes a method for displaying review information of an item. One embodiment of the method may include the following steps: obtaining a user's connection request for a first page from a user terminal, the first page including detailed information of a target item; identifying one or more review pieces of information corresponding to the target item; extracting one or more review keywords based on the one or more review pieces of information; generating first page information for outputting the first page in such a way that at least a portion of the first page includes a review keyword area for displaying the one or more review keywords; and transmitting the first page information to the user terminal; wherein each of the one or more review pieces of information corresponding to the target item includes review score information and detailed evaluation information of the target item.

[0008] In one embodiment, the step of extracting the above-mentioned comment keyword information may include the following steps: selecting one or more candidate comment information, wherein the candidate comment information is at least a portion of one or more comment information corresponding to the above-mentioned target item; and extracting the above-mentioned one or more comment keywords based on the detailed evaluation information of the selected candidate comment information.

[0009] In one embodiment, the step of selecting the above-mentioned candidate comment information may include the following steps: selecting candidate comment information from the above-mentioned one or more comment information based on the comment score information corresponding to the above-mentioned one or more comment information.

[0010] In one embodiment, the step of selecting the above-mentioned candidate comment information may include the following steps: selecting candidate comment information from the above-mentioned one or more comment information based on the length of the detailed evaluation information corresponding to the above-mentioned one or more comment information.

[0011] In one embodiment, the step of selecting the above-mentioned candidate comment information may include the following steps: selecting comment information written by one or more customers with similar preferences as candidate comment information, wherein the comment information is identified from the above-mentioned one or more comment information based on the information of the above-mentioned users.

[0012] In one embodiment, the step of extracting one or more comment keywords may include the following steps: extracting a plurality of words included in one or more detailed evaluation information corresponding to each of the one or more comment information; calculating the occurrence frequency corresponding to each of the plurality of words; and extracting comment keywords from the plurality of words based on the occurrence frequency.

[0013] In one embodiment, the step of extracting comment keywords based on the above-mentioned frequency of occurrence may include the following steps: calculating the similarity between the plurality of words; clustering the plurality of words based on the similarity between the plurality of words; extracting one or more candidate keywords based on the clustering result; classifying the plurality of words in a manner corresponding to each of the one or more candidate keywords; calculating the frequency of occurrence of each of the one or more candidate keywords based on the frequency of occurrence of the words corresponding to each of the one or more candidate keywords; and selecting one or more comment keywords from the candidate keywords based on the frequency of occurrence of each of the candidate keywords.

[0014] In one embodiment, the step of extracting comment keywords from the plurality of words may include the following steps: setting a weighting value corresponding to each of the above-mentioned comment information; extracting comment keywords from the plurality of words based on the frequency of occurrence corresponding to each of the above-mentioned words and the weighting value set in the comment information corresponding to each of the above-mentioned words.

[0015] In one embodiment, the weighting value may be set based on the writing time of the comment information corresponding to each of the plurality of words.

[0016] In one embodiment, the weighting value may be set based on the score information of the comment information corresponding to each of the plurality of words.

[0017] In one embodiment, the step of extracting the comment keywords corresponding to the target item may include the following steps: extracting the user's interest keywords based on the user's purchase history; and extracting the one or more comment keywords from one or more comment information corresponding to the target item based on the interest keywords.

[0018] In one embodiment, the step of extracting the user's keywords of interest may include the following steps: identifying the user's comment history; and extracting the user's keywords of interest based on the user's comment history.

[0019] In one embodiment, the step of extracting the user's keywords of interest may include the following steps: identifying the user's purchase history for one or more similar items, wherein the one or more similar items correspond to the target item among one or more purchased items identified based on the purchase history; identifying the user's review history corresponding to the one or more similar items purchased by the user; and extracting the user's keywords of interest based on the user's review history corresponding to the one or more similar items.

[0020] In one embodiment, the comment keyword area may include keyword icons corresponding to each of the one or more comment keywords, and the first page may further include a comment display area, which displays comment content in at least a portion of the area corresponding to the comment keyword area. The steps for generating the first page information may include the following steps: selecting a first comment keyword from the one or more comment keywords based on user input corresponding to the keyword icon; identifying keyword-related comment information corresponding to the selected first comment keyword from the one or more comment information corresponding to the target item; and generating the first page information by displaying the keyword-related comment information corresponding to the first comment keyword in the comment display area.

[0021] In one embodiment, the step of generating the first page information by displaying the keyword-related comment information in the comment display area may include the following steps: identifying one or more keyword-related comment information in the one or more comments, including words corresponding to the selected first comment keyword; and selecting a portion of the keyword-related comment information based on at least one of the following: writing time, author credibility, and keyword relevance corresponding to each of the one or more keyword-related comments; and generating the first page information by displaying the selected portion of the keyword-related comment information in the comment display area.

[0022] In one embodiment, the step of generating the first page information by displaying the keyword-related comment information in the comment display area may include the following steps: identifying a plurality of words included in the keyword-related comment information; identifying one or more keyword-related words corresponding to the selected comment keyword among the plurality of words; and generating the first page information by displaying the keyword-related comment information in the comment display area, wherein the keyword-related comment information is obtained by giving visual effects to the identified keyword-related words.

[0023] In one embodiment, the above visual effect may include a change in the font of the text corresponding to the above-mentioned keyword-related words.

[0024] The present invention provides an electronic device for displaying review information of an article. An embodiment of the electronic device of the present invention may include: one or more processors; and one or more memories storing commands executed by the one or more processors; and the electronic device is configured such that, when executing the commands, it executes the methods of various embodiments of the present invention.

[0025] This invention provides a non-transitory computer-readable recording medium for recording commands used to display comment information about items. The commands recorded in the recording medium according to an embodiment of this invention can be commands executed by one or more processors or commands to be implemented on a computer. The corresponding commands are configured such that, when executed by one or more processors, the one or more processors execute the methods of various embodiments of this invention. [Effects of the Invention]

[0026] According to at least one embodiment of the present invention, review information that can be provided to a user can be identified for an item, and the review information and detailed information can be provided to the user together, thereby improving the user experience and inducing the user to make a purchase.

[0027] According to at least one embodiment of the present invention, by providing abbreviated comment information when providing comment information, users can obtain the information they need, and it is expected to reduce user fatigue and improve efficiency.

[0028] According to at least one embodiment of the present invention, by selectively providing review information of items, personalized information can be provided to each user, which is expected to improve the user experience and improve the efficiency of review information transmission.

[0029] The effects based on the technical concept of the present invention are not limited to the above-mentioned effects. Those skilled in the art can clearly understand other effects not mentioned based on the description in the specification.

Implementation Method

[0031] The various embodiments described in this invention are illustrative of the technical concept of the invention and are not intended to limit them to specific implementation methods. The technical concept of the invention includes various modifications, equivalents, alternatives, and embodiments obtained by selectively combining all or part of the embodiments described in this invention. Furthermore, the scope of the technical concept of the invention is not limited to the various embodiments presented below or their specific descriptions.

[0032] The terms used in this invention, including technical or scientific terms, shall have the meanings commonly understood by those with common knowledge in the technical field to which this invention pertains, unless otherwise defined.

[0033] The expressions such as "comprising," "may include," "possibly possess," "may possess," "have," and "may have" used in this invention mean that features that are part of the object (e.g., functions, actions, or constituent elements) exist, and do not exclude the existence of other additional features. That is, the expressions described above should be understood as open-ended terms that have the possibility of including other embodiments.

[0034] Regarding the singular expressions used in this invention, unless otherwise stated in the context, they may include the meaning of plural expressions, and this also applies to the singular expressions described in the claims.

[0035] The terms “first,” “second,” “first,” “second,” etc., used in this specification are used to distinguish one object from another when referring to multiple objects of the same kind, unless the context otherwise requires, and are not used to limit the order or importance of such objects.

[0036] The expressions “A, B and C”, “A, B or C”, “A, B and / or C”, “at least one of A, B and C”, “at least one of A, B or C”, “at least one of A, B and / or C”, etc., used in this invention may refer to each of the listed items or all possible combinations of the listed items. For example, “at least one of A or B” may refer to (1) at least one A, (2) at least one B, (3) at least one A and at least one B.

[0037] The term "based on" used in this invention is used to describe one or more factors that affect the determination or judgment behavior or action described in the statement or article containing the term. The term does not exclude other factors that affect the determination or judgment behavior or action.

[0038] The expression that a certain constituent element (e.g., the first constituent element) is "connected" or "connected" to another constituent element (e.g., the second constituent element) in this invention may refer not only to the fact that the aforementioned constituent element is directly connected to or connected to the aforementioned other constituent element, but may also refer to the fact that it is connected to or connected to the aforementioned other constituent element by a new other constituent element (e.g., the third constituent element).

[0039] The expression "configured to" used in this invention may have meanings such as "configured in a ~ manner," "possessing the ~ capability," "modified in a ~ manner," "manufactured in a ~ manner," and "capable of ~" depending on the context. This expression is not limited to the meaning of "specially designed in hardware." For example, a processor configured to perform a specific action may refer to a general-purpose processor that can perform that specific action by executing software, or a special-purpose computer that performs that specific action by programming.

[0040] Hereinafter, various embodiments described in the present invention will be described with reference to the accompanying drawings. In the drawings and descriptions, the same or substantially equivalent constituent elements are given the same reference numerals. Furthermore, in the following descriptions of embodiments, repeated descriptions of the same or corresponding constituent elements may be omitted; however, this does not imply that the constituent element is not included in the embodiment.

[0041] Figure 1 illustrates an environment in which an embodiment of the present invention can be applied. In one embodiment, electronic device 101 and user terminal 102 can be connected and communicate with each other via a network. Electronic device 101 can be an electronic device for providing e-commerce services. That is, electronic device 101 can be a device operated under the management of an e-commerce service operator (e.g., an operator).

[0042] In one embodiment, the electronic device 101 can manage information about users using e-commerce services. The user information managed by the electronic device 101 may include, for example, user name information, user account information, age information, address information, or contact information. The management of user information can refer to all actions involving control over and processing of the relevant information. For example, the management of user information may include actions such as storing, updating, or modifying the relevant information.

[0043] In one embodiment, the electronic device 101 can manage information about items in an e-commerce service. An e-commerce service may refer to a service that provides users with the ability to purchase items such as goods, food, or products; furthermore, it may include a service that mediates the sale of items from one group of users to another. As described above, the information about items managed by the electronic device 101 may include, for example, item identification information, item name, item information, item price, item capacity, item size, item weight, item attributes, category, or unit price information. In this case, an item may be a sales unit of a product, and even in the case of the same product, it may be provided to the user as the same item. For example, even in the case of the same product, it may be sold as different items using different sales unit quantities and prices. Therefore, there may be identical items sharing product information. The management of item information can refer to all actions involving having control over and processing the relevant information. For example, the management of information about items may include actions such as storing, updating, or correcting the relevant information.

[0044] In one embodiment, electronic device 101 can manage various information to be provided to users using e-commerce services. Electronic device 101 can manage various data required to provide e-commerce services based on requests from operators of e-commerce services or users using e-commerce services. For example, it can generate, modify, and store data for various pages accompanying the e-commerce service, namely, data for search results pages, item catalog pages, event pages, item description pages (or product description pages), item sales pages, and login pages corresponding to user access requests, or provide them to the user. Regarding the management method of electronic device 101, electronic device 101 can manage the information provided accompanying the e-commerce service based on user input.

[0045] In one embodiment, the electronic device 101 may provide page information to user terminal 102 for displaying pages related to the e-commerce service based on a request from a user using an e-commerce service. The user terminal 102 may output the page on the screen based on page information received from the electronic device 101 . Among them, the pages may include: pages related to the user's login to the e-commerce service, a page that provides various information about items that the user can purchase or an item catalog page that provides a catalog of items retrieved in the various ways requested by the user, a shopping cart page used to purchase items packed in a scheduled purchase list (example: a shopping cart), an item detailing page (also a product detailing page) that provides details of a single item, a sale page of items used to purchase a single item, etc. The page feed action may include an action that will provide to the user terminal 102 at least a portion of all data generated in the course of the user's use of the e-commerce service.

[0046] In one embodiment, the electronic device 101 may manage transactions related to an e-commerce service provided to a user, i.e., transactions related to items on the line (example: goods or services). That is, the electronic device 101 may manage a range of processes such as a user can retrieve, select, purchase and settle items and request distribution. As such processing, the electronic device 101 may provide the customer with information about the items and distributions and receive purchase, settlement and distribution requests from the user from the user so that the user may retrieve, select, purchase and settle the items and request distribution. In addition, it is noted that with respect to an action that can be implemented by electronic devices 101, even if it is not mentioned in the invention, as long as it is a usual action that can be implemented by a server providing a well-known e-commerce service, the application of the technical ideas of the invention with reference to that action is not excluded within the scope of the invention.

[0047] The above electronic device 101 can be implemented with one or more computing servers. For example, all functions of the electronic device 101 may be implemented in a single computing server. For further example, the 1st function of the electronic device 101 may be implemented in the 1st computation server and the 2nd function may be implemented in the 2nd computation server. As yet another example, a plurality of computing servers that implement all functions or a particular function of the electronic device 101 , respectively, can be used. The above computing servers may be, but are not limited to, Desktop Computers, Laptop Computers, Application Servers, Proxy Servers or Cloud Servers.

[0048] In one embodiment, the user terminal 102 may be a device used by a user to access e-commerce services. The user terminal 102 may be a terminal that sends and receives various information with the electronic device 101 via a network. For example, the user terminal 102 may be one of a desktop computer, a laptop computer, a portable communication terminal (smartphone, etc.), a portable multimedia server, a wearable server, or an HMD (Head-Mounted Display). However, the type of user terminal 102 is not limited to these. The user terminal 102 may be any server that includes an input / output interface that can receive information from or output information to the user, and communicates with the electronic device 101 or other servers via a network.

[0049] In one embodiment, the user terminal 102 can provide information received from the electronic device 101 to the user, and can receive input from the user and transmit it to the electronic device 101. Specifically, the user terminal 102 can obtain input from the user instructing to call various web pages, and generate commands instructing to call various web pages corresponding to the obtained input. The user terminal 102 can transmit commands instructing to call various web pages to the electronic device 101. The input obtained from the user may include various forms of input such as clicking using a mouse, touching using a touchpad or touchscreen, voice recognition, and other electronic inputs. The user terminal 102 can receive information from various web pages from the electronic device 101, and output web pages according to the received information from various web pages.

[0050] In one embodiment, the network can serve to connect the electronic device 101 to the user terminal 102 or other external servers. For example, the network can provide a connection path so that the user terminal 102 or other external servers can connect to the electronic device 101 to send and receive packet data with the electronic device 101. The network can be implemented as any type of wired or wireless network, such as a Local Area Network (LAN), a Wide Area Network (WAN), a Mobile Radio Communication Network, or Wibro (Wireless Broadband Internet).

[0051] In one embodiment, the electronic device 101 and the user terminal 102 can operate as a single device. The user terminal 102 can be included in the electronic device 101 as a component of all or part of the electronic device 101. In this case, for example, the various information sent and received between the electronic device 101 and the user terminal 102 via a network can be various information sent and received between components within a server.

[0052] Figure 2 is a block diagram of a server according to an embodiment of the present invention. In one embodiment, electronic device 200 (e.g., electronic device 101 in Figure 1) can process information for e-commerce services. Electronic device 200 may include one or more processors 210, one or more memory units 220, and communication interface 230 as constituent elements. At least one of the constituent elements of electronic device 200 may be omitted, or other constituent elements may be added to electronic device 200. In one embodiment, some constituent elements may be additionally or alternatively integrated, or implemented as a single or multiple individual elements. In the present invention, one or more processors 210 may be referred to as processor 210. Unless explicitly stated otherwise in the context, the term processor 210 may refer to a collection of one or more processors. In the present invention, one or more memory units 220 may be referred to as memory units 220. Unless explicitly stated otherwise in the context, the term "memory 220" may refer to a collection of one or more memories. In one embodiment, at least some of the internal / external components of the electronic device 200 are interconnected via buses, GPIO (General Purpose Input / Output), SPI (Serial Peripheral Interface), or MIPI (Mobile Industry Processor Interface), thereby enabling the transmission and reception of information (data, signals, etc.).

[0053] In one embodiment, the processor 210 of the electronic device 200 can drive software (e.g., commands, programs, etc.) to control at least one component of the electronic device 200 connected to the processor 210. Furthermore, the processor 210 can perform various operations related to the present invention, such as calculations, processing, data generation, and manipulation. Also, the processor 210 can load data from the memory 220 or store data in the memory 220. The processor 210 can control the communication interface 230 to request various information from the user terminal 102 and receive various information from the user terminal 102.

[0054] In one embodiment, the memory 220 of the electronic device 200 may store various information (data). The information stored in the memory 220 is information obtained, processed, or used by at least one component of the electronic device 200, and may include software (e.g., commands, programs, etc.). The memory 220 may include volatile and / or non-volatile memory. In this disclosure, commands or programs are software stored in the memory 220, and may include an operating system for controlling the resources of the electronic device 200, application software, and / or middleware that provides various functions to the application software so that the application software can utilize the resources of the electronic device 200. In one embodiment, the memory 220 may store commands that cause the processor 210 to perform operations when executed by the processor 210. The memory 220 may store at least a portion of information received from a database through the communication interface 230 and / or information sent to a database through the communication interface 230. Specifically, memory 220 can store information about e-commerce services and commands executed by processor 210.

[0055] In one embodiment, the communication interface 230 can perform wireless or wired communication between the electronic device 200 and a database or other external device. For example, the communication interface 230 can implement wireless communication based on eMBB (enhanced Mobile Broadband), URLLC (Ultra Reliable Low-Latency Communications), MMTC (Massive Machine Type Communications), LTE (Long-Term Evolution), LTE-A (LTE Advance), NR (New Radio), UMTS (Universal Mobile Telecommunications System), GSM (Global System for Mobile communications), CDMA (Code Division Multiple Access), WCDMA (Wideband CDMA), WiBro (Wireless Broadband), WiFi (Wireless Fidelity), Bluetooth, NFC (Near Field Communication), GPS (Global Positioning System), or GNSS (Global Navigation Satellite System). For example, the communication interface 230 can implement wired communication based on USB (universal serial bus), HDMI (high definition multimedia interface), RS-232 (recommended standard 232), or POTS (plain old telephone service). In one embodiment, the electronic device 200 can also be integrated with other servers. In this case, the communication interface 230 can function as a connection circuit or interface connecting the electronic device 200 to corresponding other servers.

[0056] A user terminal according to an embodiment of the present invention (e.g., user terminal 102 in FIG1) may also include one or more processors, one or more memory units, and a communication interface as constituent elements. At least one of the constituent elements of the user terminal may be omitted, or other constituent elements may be added to the user terminal. In one embodiment, some constituent elements may be additionally or alternatively integrated, or implemented as a single or multiple individuals. One or more processors of the user terminal may drive software (e.g., commands, programs, etc.) to control at least one constituent element of the user terminal connected to the processor. One or more memory units of the user terminal may store various information (data). The communication interface of the user terminal may enable wireless or wired communication between the user terminal and a database or other external devices.

[0057] Figure 3 is a flowchart illustrating a method for displaying comment information of an item according to an embodiment of the present invention. The method for displaying comment information of an item according to an embodiment of the present invention may include actions S310, S320, S330, and / or S340.

[0058] Electronic device 101 can obtain the user's connection request for the first page (S310).

[0059] In one embodiment, the first page may be any of a variety of pages provided in conjunction with the e-commerce service. For example, the electronic device 101 may receive a user's request for a link to a detailed description page of an item in the e-commerce service.

[0060] In one embodiment, the user terminal 102 may transmit a connection request to the electronic device 101 for a first page corresponding to a target item based on user input. The user input refers to input from the user to the user terminal 102, for example, it may represent the user touching or tapping at least a portion of the area displayed for the items included in the item list page.

[0061] The electronic device 101 can identify comment information corresponding to the target item (S320).

[0062] In one embodiment, the electronic device 101 may identify comment information corresponding to the target item in response to a connection request received from the user terminal 102.

[0063] Review information may include reviews and ratings written by one or more customers for the target item. For example, review information may include at least review score information. Score data may include data on the scores given by customers (including users) for the item. The customer of the item may refer to a customer who purchases the e-commerce service for the item. Review score information may be data that quantifies the customer's evaluation of the item. For example, review score information may be quantified as a real number above 0 and below 5.

[0064] The review information may include detailed evaluation information registered corresponding to the review scores given by one or more customers for the target item. For example, the review information may include detailed evaluation information written corresponding to the review scores. The detailed evaluation information may be written corresponding to the review scores, and the correspondence between the review scores and the detailed evaluation information may be stored in the electronic device 101. Exemplarily, the review information may include at least the review scores, or may not include the detailed evaluation information. When the detailed evaluation information is written corresponding to the review scores, the review information may include the review scores and the corresponding detailed evaluation information. The detailed evaluation information may, for example, be data written and registered by the seller of the item, the administrator of the e-commerce service, etc., corresponding to the review scores. The detailed evaluation information may include at least a portion of text, images, and audio.

[0065] In one embodiment, the review information may be data written, registered, and stored in relation to an item. For example, review score information assigned to a specific item and detailed evaluation information written in relation to the review score information may be included in the review information. The electronic device 101 may identify one or more review pieces of information corresponding to one or more items included in the item catalog on the first page.

[0066] In one embodiment, the electronic device 101 may identify comment information based on item information. For example, the electronic device 101 may identify one or more comment information stored corresponding to item identification information.

[0067] In one embodiment, the electronic device 101 may identify comment information based on item information. For example, the electronic device 101 may identify one or more comment information stored corresponding to item identification information.

[0068] In one embodiment, the electronic device 101 can manage one or more review information corresponding to an item according to review category. For example, when a user writes review score information, category information corresponding to the review score information can be entered at the same time. The category can be different for each item based on the item information. For example, when the item is food, the category used to classify the review information can be classified according to freshness, taste, volume, etc., and when the item is clothing, the corresponding category can be classified according to size accuracy, color accuracy, quality, etc.

[0069] The electronic device 101 can extract comment keywords based on comment information (S330).

[0070] In one embodiment, the electronic device 101 can parse detailed evaluation information to extract one or more keywords.

[0071] In one embodiment, the electronic device 101 can parse detailed review information to extract one or more words included in the detailed review information, and extract review keywords corresponding to a portion of each word. Review keywords can be extracted from one or more review messages corresponding to the target item. Review keywords can refer to keywords that imply the meaning of the review information corresponding to the target item. There can be one or more review keywords corresponding to the target item. Review keywords do not necessarily have to be the same as the words written in the detailed review information of the review information; they can be words that cover the same or similar meanings or words that have been written.

[0072] In one embodiment, the electronic device 101 may extract comment keywords based on the frequency of occurrence, which corresponds to a plurality of words extracted based on detailed review information.

[0073] In one embodiment, the electronic device 101 may extract comment keywords based on the result of clustering the extracted plurality of words. For example, the electronic device 101 may cluster the plurality of words based on the similarity between the plurality of words.

[0074] In one embodiment, the electronic device 101 can extract one or more candidate keywords based on the clustering results. For example, words with high similarity to each other can be clustered into a cluster, and the word with the highest similarity to other words within the corresponding cluster can be equivalent to a representative word, i.e., a keyword. At this time, the representative words of each cluster can be identified as candidate keywords. The electronic device 101 can calculate the frequency of occurrence of the words corresponding to each candidate keyword, i.e., the frequency of occurrence of the words included in each cluster. The electronic device 101 can select comment keywords from one or more candidate keywords based on the frequency of occurrence of each candidate keyword, i.e., the frequency of occurrence of the words included in each cluster.

[0075] In one embodiment, the electronic device 101 can set a weighting value corresponding to each comment piece of information. The weighting value can be determined based on the writing time of each comment piece of information, the credibility of the writer, the relevance of each comment piece of information to the item, and whether the comment piece of information includes keywords that are highly sensitive to the user. The electronic device 101 can extract comment keywords based on the weighting values ​​corresponding to each comment piece of information. For example, when extracting comment keywords based on the frequency of occurrence of each word, the weighting value of the comment piece of information corresponding to each word can be applied to each frequency of occurrence. The weighting value corresponding to the first comment piece of information and the weighting value corresponding to the second comment piece of information can be applied to the frequency of occurrence of each of the words A included in the first comment piece of information and the words B included in the second comment piece of information.

[0076] In one embodiment, the electronic device 101 can extract keywords from detailed evaluation information based on Term Frequency-Inverse Document Frequency (TF-IDF) values. The electronic device 101 can select keywords from one or more keywords extracted through parsing, based on words that frequently appear in each detailed evaluation piece of information and words that are rare in the entire document. The electronic device 101 can calculate the Term Frequency (TF) of a specific word within the review score information and calculate the Inverse Document Frequency (IDF) that takes into account the frequency of the corresponding word in the entire set of review score information. The electronic device 101 can multiply the TF and IDF values ​​to determine the importance of each word.

[0077] In one embodiment, the electronic device 101 embeds one or more keywords extracted from the parsing into a vector space, calculates the similarity between the keywords, and clusters keywords with similar meanings. The electronic device 101 can thereby extract key keywords from each detailed evaluation information.

[0078] In one embodiment, the electronic device 101 can set words in the detailed evaluation information as nodes and the connections between words as edges to calculate the importance of each word. The electronic device 101 can extract comment keywords from the comment information based on the importance of each word.

[0079] In one embodiment, the electronic device 101 can use various methods to calculate distance information between keywords and cluster detailed evaluation information accordingly. For example, the electronic device 101 can use the cosine similarity method to convert each keyword included in the review score information into a vector and then calculate the cosine similarity to calculate the similarity between keywords. Cosine similarity measures the angle between two vectors, so the closer the value is to 1, the more similar the two keywords are considered to be. The electronic device 101 can cluster more than one piece of detailed evaluation information based on the similarity between keywords.

[0080] In one embodiment, the electronic device 101 can convert keywords into numerical vectors using Euclidean distance, and then calculate a numerical vector including comment score information for each keyword. Furthermore, the electronic device 101 can calculate the Euclidean distance between vectors. The smaller the Euclidean distance value, the higher the similarity between keywords can be determined, and clustering can be performed accordingly.

[0081] In one embodiment, the electronic device 101 may use Jaccard Similarity to calculate the Jaccard similarity between keyword sets to measure the similarity between keywords. Jaccard Similarity is defined as the value obtained by dividing the size of the intersection of two sets by the size of the union of the two sets; therefore, the higher the similarity, the larger the value. This allows for the clustering of detailed evaluation information.

[0082] In one embodiment, the electronic device 101 may utilize word embedding, such as Word2Vec or GloVe, to map keywords to a vector space. Then, the similarity between keywords can be measured using cosine similarity or Euclidean distance between vectors, and clustering of detailed evaluation information for each keyword can be performed accordingly.

[0083] In one embodiment, the electronic device 101 may use TF-IDF (Term Frequency-Inverse Document Frequency) values ​​to construct document vectors for each detailed evaluation information. Then, cosine similarity or Euclidean distance is used to calculate the distance between document vectors, and the detailed evaluation information is clustered.

[0084] In one embodiment, the electronic device 101 can extract main keywords for clusters of clustered comment information and define the category of each cluster based on the main keyword information.

[0085] In one embodiment, electronic device 101 may map and store category information corresponding to each comment information.

[0086] In one embodiment, the electronic device 101 can extract one or more comment keyword information corresponding to the target item based on the user's interest keywords.

[0087] In one embodiment, the electronic device 101 may extract keywords of interest based on the user's purchase history.

[0088] In one embodiment, the electronic device 101 can extract keywords of interest based on a user's purchase history. In particular, it can consider not only the user's purchase history but also the user's previous comments. To this end, the electronic device 101 can grasp the user's past purchase history information and comment writing history, and identify the user's keywords of interest based on the corresponding purchase and comment data.

[0089] Electronic device 101 can analyze the history of products purchased by users to identify frequently purchased product groups. Then, it can analyze user reviews within these product groups to extract keywords, including features that are repeatedly expressed or emphasized. For example, if words such as "battery life" and "screen resolution" frequently appear in reviews of electronic devices that users frequently purchase, these words can be identified as keywords of interest to the user.

[0090] In one embodiment, the electronic device 101 can search for similar items with a high degree of relevance to the target item based on the user's purchase history and analyze the user's comment history on the corresponding items. This is a process of confirming whether the user shows a consistent preference for a specific category or attribute, thereby deriving more accurate keywords of interest. For example, the keywords frequently commented by the user in a group of smartphone products purchased multiple times can be analyzed to understand the user's preferences.

[0091] After analyzing user comment patterns based on the comment history of one or more similar items, the core words that repeatedly appear in the corresponding patterns are extracted and identified as keywords of interest. These keywords of interest can be used to more accurately parse comment information corresponding to the target item (e.g., smartphones, clothing, home appliances, etc.) and provide the extracted comment keywords in a user-customized format.

[0092] Furthermore, the electronic device 101 can extract comment keywords based on the keywords extracted by the TF-IDF method, and based on the similarity to the keywords of interest calculated using various distance calculation methods such as cosine similarity, Euclidean distance, and Jaccard similarity. Using this method, comment keywords can be more accurately clustered, and comment keywords that match the user's keywords of interest can be identified in the corresponding clusters.

[0093] Therefore, in the process of extracting comment keywords from comment information corresponding to the target item, the electronic device 101 derives comment keywords by utilizing interest keywords identified based on the user's purchase history and comment history, thereby providing the user with customized comment information. This allows for the provision of more reliable and meaningful comment keywords to the user.

[0094] In one embodiment, the electronic device 101 may extract comment keywords based on a portion of the comment information from a plurality of comment information corresponding to the target item.

[0095] In one embodiment, the electronic device 101 may select a portion of the comment information, namely, candidate comment information, from a plurality of comment information corresponding to the target item to extract one of the comment keywords.

[0096] In one embodiment, the electronic device 101 can select candidate comment information from one or more comment information based on comment score information corresponding to each comment information. For example, the electronic device 101 can select comment information with a comment score above a specified threshold score as candidate comment information.

[0097] In one embodiment, the electronic device 101 may select candidate comment information based on the length of the detailed evaluation information corresponding to the comment information. The electronic device 101 may select comment information whose detailed evaluation information has a text length of a specified first length or greater as candidate comment information. The electronic device 101 may also select comment information whose detailed evaluation information has a text length of a specified second length or less as candidate comment information. In this case, the first length is set to be shorter than the second length.

[0098] In one embodiment, the electronic device 101 may select candidate comment information based on the time when the comment information was written. The electronic device 101 may select comment information written within a specified period as candidate comment information.

[0099] In one embodiment, the electronic device 101 may exclude comments written before a specified critical period from one or more comment information as the basis for selecting comment keywords, i.e., from candidate comment information. The electronic device 101 may exclude comments from the basis for selecting comment keywords based on at least a portion of the writing time corresponding to each comment information, the relevance between the items corresponding to the comment information, and the credibility of the writer.

[0100] In one embodiment, the electronic device 101 can set a weighting value for each of more than one comment piece of information. The electronic device 101 can set a weighting value corresponding to each comment piece of information based on at least one of the following: the writing time corresponding to each comment piece of information, the relevance between the items corresponding to the comment pieces of information, and the credibility of the writer. The electronic device 101 can correct the number of comments calculated by applying weighting values ​​to each comment piece of information, or correct the comment score.

[0101] In one embodiment, the electronic device 101 can analyze user profiles or query records, select the most suitable comment information for the user, and select candidate comment information based on the selected comment information. For example, candidate comment information can be selected based on the score and / or number of comment information, wherein the comment information is selected based on the similarity to comment information of items previously retrieved by the user or items of similar categories.

[0102] In one embodiment, the electronic device 101 can analyze past comment information written by the user and select comment information with a similarity of more than a critical similarity to the user's past comment information as candidate comment information.

[0103] In one embodiment, the sensitivity of each keyword attribute of a user can be calculated by considering the user's or customer's past review scores and purchase history, and a weighted value is determined for each extracted keyword based on the sensitivity. Then, based on the weighted value and the extracted keywords, the priority corresponding to each of one or more review messages is determined. In one embodiment, the electronic device 101 can analyze the user's preferred topics or keywords based on the user's past review information and purchase history. Then, based on this information, the sensitivity of specific keywords is calculated, and a weighted value is assigned to each keyword to determine the priority of the review messages. For example, in the case of a user who frequently buys athletic shoes, review messages including the keyword "size" can be set to a higher priority. Furthermore, the electronic device 101 can select a portion of the review messages based on the priority of each review message as candidate review messages.

[0104] In one embodiment, the electronic device 101 may exclude comment information that is less relevant to the target item from the comment information that forms the basis for selecting comment keywords. For example, the electronic device 101 may exclude at least a portion of comment information that is less relevant to the item from the comment information that forms the basis for selecting comment keywords.

[0105] The electronic device 101 can calculate the correlation between the item and one or more identified commenters.

[0106] For example, electronic device 101 can calculate the relevance of one or more reviews corresponding to a target item. Electronic device 101 can analyze keywords, sentence structure, and writing patterns of detailed evaluation information included in each review. In this case, the information that becomes the object of analysis by electronic device 101 can be the detailed evaluation information included in the review information. For example, when keywords such as "quality," "design," and "durability" frequently appear in the detailed evaluation information, these keywords are more likely to be closely related to the main characteristics of the corresponding item.

[0107] In one embodiment, the electronic device 101 may consider the category of the target item and calculate the correlation between the review information and the item. For example, when calculating the correlation between the review information and the target item by comparing it with customized category information, the electronic device 101 may determine that reviews of soccer shoes that include specific terms such as "grass", "cushioning" and "durability" are more correlated with the item.

[0108] In one embodiment, the electronic device 101 may assign a score to each comment. The electronic device 101 may calculate the relevance to the target item based on the assigned score. For example, when the comment on the target item "A" includes the statement "wide range of color choices," this comment fully reflects the characteristics of item "A" and is therefore evaluated as having a high relevance. Conversely, information such as "slow delivery" and "simple packaging" has a lower direct relevance to the target item "A" and may be evaluated as having a lower relevance.

[0109] Furthermore, the electronic device 101 can apply weighted values ​​to more finely adjust the relevance of comments. The credibility of the comment writer or the time of writing can be considered, assigning higher weighted values ​​to the most recent comments or comments from writers with higher credibility. For example, a comment that receives a score of 4.5 out of 5 in customer ratings can have a higher weighted value than a comment that receives a score of 3.0.

[0110] The electronic device 101 may select at least a portion of the comment information from a plurality of comment information corresponding to the target item as candidate comment information based on the credibility of the writer.

[0111] In one embodiment, the electronic device 101 may exclude comment information whose author's credibility does not reach a specified value from the comment information that serves as the basis for selecting comment keywords. For example, the electronic device 101 may exclude at least a portion of comment information whose author's credibility is low from the comment information that serves as the basis for selecting comment keywords.

[0112] The electronic device 101 can calculate the credibility of the writer of each of the more than one comments that correspond to the target item.

[0113] In one embodiment, the electronic device 101 can calculate the credibility of the author of each of the identified one or more commenters. The electronic device 101 can select at least a portion of the comment information based on the author's credibility.

[0114] In one embodiment, credibility can be determined based on the degree of relevance between the author's past comments and the items, and the evaluations of the past comments obtained from other users.

[0115] In one embodiment, in order to evaluate the relevance to a target item, the electronic device 101 may perform keyword similarity analysis. For example, it may extract keywords included in each comment and evaluate the degree of consistency between the corresponding keywords and the item attributes.

[0116] The electronic device 101 can identify comment information written in the past by the author corresponding to the comment information. When the correlation between the identified comment information and the corresponding item is low, the credibility of the corresponding author is calculated to be low.

[0117] In one embodiment, to determine the credibility of the author, the electronic device 101 may construct a multi-user rating system. The system may collect ratings such as "like" or "helpful" left by other users for the corresponding comment to reflect the author's credibility.

[0118] Furthermore, the electronic device 101 can also use the frequency of comment writing as a factor in determining credibility. For example, when the rate of increase in a writer's comment writing frequency exceeds a critical value, there is a possibility that the writer's comments are spam, and thus the writer can be assigned a lower credibility rating. Conversely, users who consistently write comments in a consistent pattern can receive a higher credibility rating.

[0119] In one embodiment, electronic device 101 can analyze a reviewer's past purchase history and review patterns to calculate the reviewer's credibility. For example, electronic device 101 can identify the frequency and quality of reviews written by each reviewer, categorized by item type. Quality can be identified based on at least a portion of factors, including whether the review score deviation is below a specified value, other users' evaluations of the review, and the relevance of each review to the item. A higher overall credibility rating can be obtained when a particular reviewer has a history of writing fair and detailed reviews across various categories. For example, when electronic device 101 obtains data indicating that a particular user consistently writes high-quality reviews for multiple categories (sneakers, milk, home appliances, etc.), it can assign a high credibility rating to the user's reviews.

[0120] Electronic device 101 can combine more than one comment information.

[0121] In one embodiment, the electronic device 101 may combine at least a portion of one or more comment information based on the similarity of the comment information.

[0122] In one embodiment, the similarity between each comment information can be determined based on the similarity between each detailed evaluation information included in each comment information.

[0123] In one embodiment, the electronic device 101 may combine at least a portion of one or more comment information based on the similarity of the comment information. For example, the electronic device 101 may analyze the detailed evaluation information of each comment information and combine data including similar detailed evaluation information.

[0124] In one embodiment, the electronic device 101 can combine multiple pieces of review information based on the similarity between scores included in each piece of review information and the similarity between detailed evaluation information. For example, the electronic device 101 can identify multiple pieces of review information with similar detailed evaluation information, and can combine such multiple pieces of similar review information when the difference between the review scores of the identified multiple similar review information is below a critical score. In this case, the method for determining whether the difference between review scores is below the critical score can be, for example, based on the average review score of the multiple similar review information and the deviation of each piece of review information from the average review score. That is, review information whose review score deviates from the average review score beyond a critical deviation can be excluded from the review information to be integrated.

[0125] In one embodiment, the electronic device 101 may construct document vectors to determine similarity. For this purpose, the electronic device 101 may use TF-IDF (Term Frequency-Inverse Document Frequency) values ​​to construct document vectors for each comment piece of information. Then, the cosine similarity or Euclidean distance between the document vectors may be calculated to measure the similarity between the detailed evaluation information included in each comment piece of information.

[0126] In one embodiment, the electronic device 101 may use word embedding. The electronic device 101 may use word embedding methods such as Word2Vec and GloVe to map the review score information to a vector space, and then calculate the cosine similarity or Euclidean distance between the vectors to measure the similarity of the detailed review information.

[0127] In one embodiment, the electronic device 101 may use a language analysis model to convert the contextual description of each comment information into a vector and calculate the cosine similarity between the transformed vectors to measure the similarity between detailed evaluation information.

[0128] Electronic device 101 can merge review information, including similar detailed evaluation information, into a single file. Similar review information may refer to multiple detailed evaluation information pieces with a similarity of more than a specified critical similarity. For example, review information containing similar content such as "easy to take" and "this product is easy to use" can be combined.

[0129] In one embodiment, the electronic device 101 can associate an integrated set of detailed evaluation information with each of the multiple sets of comment information, which includes multiple sets of similar detailed evaluation information, while keeping each set of comment score information unchanged. In one embodiment, when the multiple sets of comment information, which includes multiple sets of similar comment score information, include detailed evaluation information, the electronic device 101 can combine the detailed evaluation information into one. For example, the electronic device 101 can calculate the similarity for each set of detailed evaluation information corresponding to the similar comment score information, and extract representative detailed evaluation information based on the similarity. For example, the electronic device 101 can calculate the similarity between each pair of detailed evaluation information for each of the three sets of detailed evaluation information corresponding to each of the three sets of similar comment score information, and based on this, identify the detailed evaluation information that becomes the center among the detailed evaluation information. In this scenario, for example, if there are three sets of review information and the similarity between the detailed evaluation information and the detailed evaluation information is above a specified level, only the detailed evaluation information can be integrated while the individual scores of the three reviews remain unchanged. In this case, the number of review information is identified as three, and only the detailed evaluation information is integrated into one.

[0130] In one embodiment, the electronic device 101 may use various methods such as cosine similarity, Jaccard similarity, or Euclidean distance to calculate similarity. For example, when the electronic device 101 calculates the cosine similarity for three detailed evaluation information pairs A, B, and C, after calculating the cosine similarity between each pair of detailed evaluation information pairs (AB, AC, BC), the two detailed evaluation information pairs with the highest similarity can be identified. Subsequently, representative detailed evaluation information can be extracted using a weighted average or majority voting method based on the identified pairs of detailed evaluation information pairs. In this method, the electronic device 101 can quantitatively evaluate the content similarity of each detailed evaluation information pair, maintaining consistency in detailed evaluation information for similar reviews. For example, when the cosine similarity between detailed evaluation information pairs A and B is 0.9, the similarity between A and C is 0.85, and the similarity between B and C is 0.8, the electronic device 101 can determine that A and B have the highest similarity and combine the two to generate representative detailed evaluation information. In one embodiment, the electronic device 101 can combine similar comment information and store the combined data as new comment information. This combined comment information contains more comprehensive information, improving the quality of information provided to the user.

[0131] In one embodiment, the electronic device 101 can classify review information corresponding to an item into one or more categories. The electronic device 101 can generate one or more categories for classifying the one or more review information based on the analysis of the review information. For example, the electronic device 101 can analyze one or more detailed review information included in one or more review information sets, and generate one or more categories for classifying each detailed review information set based on the analysis. In one embodiment, the electronic device 101 can extract keywords from one or more detailed review information sets, calculate distance information between keywords included in each review score information set, and cluster the detailed review information sets into one or more clusters based on the calculated distance information.

[0132] In one embodiment, the electronic device 101 may store information on pre-specified review keywords for a product.

[0133] In one embodiment, the electronic device 101 may apply visual effects (e.g., highlighting) to words in the comment information that correspond to the comment keywords. For example, the electronic device 101 may apply visual effects to words that correspond to the comment keywords among a plurality of words in the detailed review information included in the comment information and display the comment information. This allows the user to quickly identify important information. Words corresponding to comment keywords may, for example, refer to words that are the same as or similar to the comment keywords.

[0134] In one embodiment, the electronic device 101 may apply visual effects to words that match a pre-specified list of comment keywords.

[0135] In one embodiment, electronic device 101 may extract new comment keywords or map them to existing comment keywords based on the combined comment information. The resulting comment keywords make it easier for users to find relevant information. Through the various methods illustrated in the examples above, electronic device 101 can combine comment information and provide users with more useful and comprehensive information.

[0136] In one embodiment, the electronic device 101 can delete comment information containing inappropriate expressions or defamatory language. The electronic device 101 can analyze the text of detailed evaluation information and delete data containing defamatory language, hate speech, or other inappropriate expressions. For example, the electronic device 101 can pre-store data containing defamatory language, hate speech, or other inappropriate expressions, and based on this, score the text of each piece of data according to the similarity to each word or phrase included in the detailed evaluation information, classifying comment information with scores above a predetermined threshold as deletion targets. In one embodiment, after filtering comment information containing inappropriate expressions, the electronic device 101 can construct candidate comment information based on the remaining data. This prevents accidental misuse and maintains the credibility of the information provided to the user. Furthermore, for detailed evaluation information containing inappropriate expressions, the electronic device 101 can delete or correct the corresponding detailed evaluation information and store it. To this end, the electronic device 101 can review the text of the detailed evaluation information, and when inappropriate expressions are found, correct the corresponding portion or delete all detailed evaluation information. As a result, electronic device 101 may include a function that automatically detects and removes inappropriate expressions to maintain the quality of comment information. In this way, the electronic device can provide users with trustworthy and useful information. For example, if a particular comment is determined to contain defamatory language, electronic device 101 can delete the corresponding data to maintain the quality of the dataset.

[0137] In one embodiment, the electronic device 101 may apply visual effects (e.g., highlighting) to words corresponding to comment keywords within detailed evaluation information. This allows the user to quickly identify important information.

[0138] Electronic device 101 may generate first page information (S340). The first page information may include information used to output the first page.

[0139] In one embodiment, the first page may be a details page that includes detailed information about the item.

[0140] In one embodiment, the first page may be a details page that includes detailed information about the item. The details page may include images, names, prices, descriptions, specifications, manufacturer information, etc. of the item.

[0141] In one embodiment, the first page may include a comment keyword area in at least a portion of its area. The comment keyword area may be an area for displaying more than one comment keyword. The comment keyword area of ​​the first page may display more than one comment keyword. For example, keyword icons corresponding to each comment keyword may be included in the comment keyword area.

[0142] In one embodiment, the electronic device 101 may select one or more comment keyword information to be displayed in the comment keyword area.

[0143] In one embodiment, the electronic device 101 can display comment information, including detailed evaluation information, in the comment display area. The comment display area can display all the content of each comment simultaneously, namely, score information and detailed evaluation information.

[0144] In one embodiment, the electronic device 101 can display comment information corresponding to each comment keyword in the comment display area based on the user's selection input for each comment keyword.

[0145] In one embodiment, the electronic device 101 can classify a plurality of comment information according to each comment keyword.

[0146] In one embodiment, the electronic device 101 may display abbreviated comment information in the comment display area. The abbreviated comment information may be comment information that includes abbreviated detailed evaluation information.

[0147] The electronic device 101 can confirm whether the length of the comment information is greater than or equal to a specified length. The electronic device 101 can confirm whether the length of the comment information is greater than or equal to a specified length based on the text length included in the detailed evaluation information within the comment information. For example, when the detailed evaluation information includes more than 100 words, or when the detailed evaluation information includes more than 300 words, the electronic device 101 can determine that the length of the comment information is greater than or equal to a specified length. In one embodiment, the electronic device 101 can confirm whether the length of the comment information is greater than or equal to a specified length based on the capacity of the comment information.

[0148] In one embodiment, electronic device 101 may generate abbreviated comment information.

[0149] Electronic device 101 can generate abbreviated review information, in which at least a portion of the detailed evaluation information in the review information is abbreviated.

[0150] In one embodiment, the electronic device 101 can abbreviate comment information to generate abbreviated comment information. The abbreviated comment information can be generated when the length of the original comment information exceeds a specified critical length. In the corresponding case, the electronic device 101 can analyze detailed evaluation information to determine the portion to be abbreviated.

[0151] In one embodiment, the electronic device 101 may extract a portion of detailed evaluation information within a predetermined length to generate abbreviated comment information. For example, the electronic device 101 may truncate the detailed evaluation information from the beginning to a specified predetermined length to generate abbreviated comment information.

[0152] In one embodiment, in addition to simply extracting text, the electronic device 101 can also abbreviate detailed review information according to semantic units. For example, the electronic device 101 can identify important parts of the detailed review information and write abbreviated review information based on the core information. If necessary, the electronic device 101 can generate abbreviated data by determining the area containing important content rather than the beginning content, and extracting a predetermined length centered on that area. In one embodiment, the electronic device 101 can use various methods other than simply extracting length when abbreviating review information. The electronic device 101 can efficiently perform abbreviation by analyzing the logical structure of each detailed review information to identify the main parts and exclude parts of lower importance.

[0153] In one embodiment, the electronic device 101 can decompose detailed evaluation information into individual sentences or semantic units, identify and reconstruct important parts within the overall context, thereby composing abbreviated comment information. In one embodiment, the electronic device 101 can utilize natural language processing algorithms to decompose text into semantic units and evaluate the importance of each unit. To this end, the electronic device 101 can grasp the theme of each sentence and select sentences that include important information within the overall context of the file to compose abbreviated comment information.

[0154] In one embodiment, the electronic device 101 may perform keyword-based analysis, identifying only statements containing important keywords from the detailed evaluation information. This allows for the generation of abbreviated comment information that retains important information from the detailed evaluation information while excluding unnecessary parts. The importance of keywords can be evaluated using benchmarks such as TF-IDF values.

[0155] In one embodiment, electronic device 101 may extract the most important statements to reconstruct them into abbreviated text. The selection criteria for important statements may include the position of each statement in the detailed evaluation information, such as the beginning, middle, or end of the file.

[0156] In one embodiment, the electronic device 101 can abbreviate detailed evaluation information by analyzing the sentence beginning and end. For example, when important content is concentrated in the beginning or end of the detailed evaluation information, abbreviation can be performed focusing on the corresponding part so that the user can effectively obtain the information. In this case, the important content can be calculated based on the importance of keywords.

[0157] In one embodiment, the electronic device 101 can structure and abbreviate detailed evaluation information. For example, text can be rearranged around important statements or keywords to generate abbreviated comment information in a structured form that includes only the core content. For example, based on keyword evaluation, statements including important keywords are identified, and the text of the detailed evaluation information is rearranged accordingly.

[0158] In one embodiment, the electronic device 101 can measure the similarity between statements, group statements with similar content, and extract representative statements from each group to generate summary statements.

[0159] In one embodiment, the electronic device 101 may generate abbreviated comment information by selecting only statements containing keywords identified as important keywords. This can be achieved through the aforementioned keyword importance evaluation, where statements containing keywords with higher importance may be included in the summary statements.

[0160] In one embodiment, the electronic device 101 can rearrange the text order of detailed evaluation information. The electronic device 101 can calculate the importance of each statement and rearrange the corresponding statements based on the importance to generate abbreviated detailed evaluation information. For example, the electronic device 101 can arrange statements with higher importance, including important information, first to generate abbreviated detailed evaluation information that truncates the rearranged detailed evaluation information to a predetermined length.

[0161] In one embodiment, the electronic device 101 can select important sentences after evaluating each sentence of detailed evaluation information, and generate abbreviated data while taking into account the connectivity between sentences. The electronic device 101 can select important sentences based on the tokenization, morpheme analysis, syntactic analysis, and semantic analysis of each sentence or the words included in each sentence.

[0162] In one embodiment, the electronic device 101 may use a keyword-based abbreviation method. The electronic device 101 may identify words or phrases corresponding to the keywords in each detailed review, and generate abbreviations only by identifying statements containing such words or phrases.

[0163] The electronic device 101 can also predefine important keywords corresponding to the categories of each review information, i.e., review keywords. To this end, the electronic device 101 can store a list of important keywords that repeatedly appear in detailed review information belonging to a specific category. This list can be structured differently according to the category of the item or the category of the review; for example, key keywords for electronic products may include "function," "battery life," and "performance." The electronic device 101 can refer to the corresponding list to evaluate the importance of the detailed review information and determine the parts to be abbreviated.

[0164] The electronic device 101 can confirm the priority among one or more comment information corresponding to the target item.

[0165] The electronic device 101 can identify the comment information to be displayed in the comment display area based on the priority of multiple comment information corresponding to the target item.

[0166] In one embodiment, after confirming the priority of a plurality of review messages corresponding to the target item, the electronic device 101 may select a portion of the review messages in descending order of priority and set them to be displayed in the review display area. For example, the review messages with higher purchase conversion rates will be displayed with higher priority in the review display area by the electronic device 101, thereby enabling users to quickly find the answers they need.

[0167] In one embodiment, the electronic device 101 can determine the priority among one or more review messages identified corresponding to a target item. For example, when each review message has been displayed in the review display area, the priority can be determined based on the rate at which customers purchase the item.

[0168] In one embodiment, the electronic device 101 can determine priority based on the rating data of one or more customers for one or more review items. For example, the rating data can be analyzed based on the number of times a customer presses the "like" button. In one embodiment, the electronic device 101 can analyze customer rating data to prioritize review items that have received high ratings. For example, the top-ranked review items can be detected based on the "like" button or 5-star rating data, and these can be displayed to the user first.

[0169] In one embodiment, the electronic device 101 may determine the priority of comment information based on comment keywords. In one embodiment, the electronic device 101 may determine the priority based on comment keywords extracted from the comment information. For example, comment information that frequently contains phrases or words corresponding to comment keywords may be judged as having high priority.

[0170] In one embodiment, the electronic device 101 can select similar items based on the item's attribute information and extract review keywords from review information corresponding to such similar items. In one embodiment, the electronic device 101 can analyze and use review information of similar items based on the item's attribute information. For example, it can extract review information of other products in similar categories and, based on keywords included in that data, prioritize displaying review information that matches the item's characteristics.

[0171] In one embodiment, the sensitivity of each keyword attribute of the user can be calculated by considering the user's or customer's past inquiry data and purchase history, and a weighted value is determined for each extracted keyword based on the sensitivity. Then, the priority corresponding to each of the more than one comment information is determined based on the weighted value and the extracted keywords. In one embodiment, the electronic device 101 can analyze the user's preferred topics or keywords based on the user's past inquiry data and purchase history. Then, the sensitivity of specific keywords is calculated based on this information, and a weighted value is assigned to each keyword to determine the priority of the comment information. For example, for users who frequently buy sneakers, comment information including the keyword "size" can be set to high priority. In one embodiment, the sensitivity of each keyword attribute can be calculated based on the user's history of past inquiry data and the user's return or refund history. For example, in the user's past comment information, the keyword attributes included in the comment information of items that the user has returned or refunded are more likely to be keywords with higher user sensitivity. Electronic device 101 can analyze review information corresponding to items with a history of past returns or refunds by the user, and extract user-sensitive review keywords based on this. Electronic device 101 can determine the priority of each review based on the extracted review keywords.

[0172] In one embodiment, the electronic device 101 can analyze customer search pattern data to determine the priority of review information. For example, it can analyze the frequency of customer search terms for specific items and determine the priority of review information based on frequently searched keywords. For example, when a large number of searches are conducted using the keyword "durability" for a specific item, review information containing the corresponding keyword can be displayed first.

[0173] In one embodiment, the electronic device 101 can analyze opinion data on social media to determine the priority of comment information. For example, it can analyze frequently mentioned keywords on social media platforms and use this to screen comment information. Comment information including keywords that are highly relevant to positive comments about a specific item can be prioritized for display.

[0174] In one embodiment, the electronic device 101 can analyze data traffic within a specific period to determine priorities. For example, priorities can be determined based on the number of connections and dwell time of pages containing specific comment information. Comment information of pages with longer dwell times and more connections can be displayed first.

[0175] In one embodiment, the electronic device 101 may determine priorities based on user feedback survey results. For example, comments rated as most useful may be displayed first based on user feedback. This allows for the provision of more useful information that reflects user opinions.

[0176] In one embodiment, the electronic device 101 can determine the priority of comment information based on response data from each device. For example, it can analyze responses between mobile devices and PC users to display optimized comment information for each device. By prioritizing frequently viewed questions on the mobile screen, the user experience can be improved.

[0177] In one embodiment, the electronic device 101 can determine priorities based on credibility evaluation data. For example, it can evaluate the credibility of review information, including seller reviews or verified user reviews, and prioritize data with higher credibility. This allows for the provision of more credible information to the user.

[0178] In one embodiment, the electronic device 101 may generate first page information in such a way that comment information selected based on priority is displayed in the comment display area.

[0179] Electronic device 101 can select at least a portion of the comment information based on priority. Electronic device 101 can select a specified number of comment information in descending order of priority.

[0180] The electronic device 101 can generate the first page information in such a way that at least a portion of the selected comment information is displayed in the comment display area.

[0181] Electronic device 101 can transmit the information on the first page to the user terminal (S350).

[0182] In one embodiment, the electronic device 101 can transmit first page information to the user terminal 102. The user terminal 102 can receive the first page information from the electronic device 101. The user terminal 102 can output the first page on the screen of the user terminal 102 based on the first page information.

[0183] Figure 4 is a diagram illustrating the first page and comment keyword area of ​​an embodiment of the present invention.

[0184] Refer to Figure 4, page 1 400 may be an item details page that includes detailed information about the item.

[0185] In one embodiment, the first page 400 may include a comment keyword area 410 in at least a portion of its area.

[0186] In one embodiment, the comment keyword area 410 may display keyword icons 401, 402, 403, 404, and 405 corresponding to one or more comment keyword information, wherein the comment keyword information is extracted by the electronic device 101 from one or more comment information corresponding to the item.

[0187] At least a portion of the comment keyword area 410 may further include a user interface for expanding the comment display area. In response to input corresponding to the user interface, the comment display area may be displayed in the area of ​​the first page 400 corresponding to the comment keyword area 410. The comment display area will be described below.

[0188] Figure 5 is a flowchart illustrating a method for displaying keyword-related comment information according to an embodiment of the present invention.

[0189] A method for displaying keyword-related comment information according to an embodiment of the present invention may include action S510, action S520, and / or action S530.

[0190] The electronic device 101 can select the first comment keyword from more than one comment keyword (S510).

[0191] In one embodiment, the electronic device 101 may select a first comment keyword based on user input corresponding to the comment keyword area. For example, the electronic device 101 may receive input to select the first comment keyword based on user input to a first keyword icon corresponding to the first comment keyword in the comment keyword area.

[0192] Electronic device 101 can identify keyword-related comment information corresponding to comment keywords (S520).

[0193] Keyword-related comment information can be selected based on the similarity between the selected first comment keyword and each comment information.

[0194] For example, electronic device 101 may prioritize identifying comment information that is highly similar to the first comment keyword selected by the user.

[0195] Electronic device 101 can filter out comment information containing a first comment keyword selected by user input. For example, all comments in the database containing the first comment keyword can be selected as the initial filtering target. In order to identify the comment information most similar to the first comment keyword among the filtered comment information, electronic device 101 can apply various similarity calculation methods. Such methods include cosine similarity, Jaccard similarity, or similarity calculation using embedding vectors. The comment information with calculated similarity can be sorted according to the similarity value, and the top n comments are determined as keyword-related comment information. At this time, the value of n can be variably set according to the system settings. In one embodiment, electronic device 101 can extract sub-keywords associated with the comment keyword, and add the superior comment information within the category as keyword-related comment information by clustering and topic modeling.

[0196] In one embodiment, the electronic device 101 can visually display the correlation between the first comment keyword and comment information.

[0197] In one embodiment, the electronic device 101 may generate the first page information in such a way that it displays keyword-related comment information selected based on at least one of the following: the writing time corresponding to the comment keyword, the credibility of the writer, and the relevance to the comment keyword, in the comment display area.

[0198] Electronic device 101 can display keyword-related comment information in the display area (S530).

[0199] In one embodiment, the electronic device 101 may generate the first page information in such a way as to display keyword-related comment information in the comment display area.

[0200] In one embodiment, the electronic device 101 may apply visual effects to comment information. For example, visual effects may be applied to words or phrases corresponding to comment keywords. Visual effects may include displaying the corresponding words or phrases in bold or different colors.

[0201] In one embodiment, the electronic device 101 can identify words or phrases associated with comment keywords from keyword-related comment information, i.e., keyword-related words. For example, the electronic device 101 can identify keyword-related words based on the results of analyzing the frequency of co-occurrence with comment keywords in the comment information. The electronic device 101 can use word embedding to map each word to a vector space and calculate cosine similarity or Euclidean distance, etc., to select words similar to keywords.

[0202] The electronic device 101 can replace keyword-related words with the corresponding comment keywords themselves, or convert keyword-related words into other similar comment keywords and display them. For example, when the comment keyword "screen" exists, it can be converted into similar keywords such as "display screen" or "screen". The electronic device 101 can provide a function that allows manual or automatic replacement. In this way, more easily understandable comment information can be provided to the user.

[0203] In one embodiment, the electronic device 101 can assign visual effects to keyword-related terms. For example, keyword-related terms can be displayed in bold for emphasis, or color coding can be applied to the corresponding words or phrases. For example, the electronic device 101 can display keyword-related terms corresponding to the comment keywords in red or blue. Furthermore, the electronic device 101 can apply a highlight effect to keyword-related terms to change the background color. Fourth, keyword-related terms can be underlined or converted to italics to assign visual effects.

[0204] In one embodiment, the electronic device 101 may use visual effects such as bold, color change, or underline to emphasize the identified keywords. For example, key keywords such as "delivery period," "refund policy," and "product features" may be displayed more prominently than other text.

[0205] In this embodiment, the electronic device 101 may provide a search function based on comment keywords highlighted according to the user's selection. For example, if the user clicks on a specific comment keyword, all comment information associated with the corresponding keyword can be displayed as search results.

[0206] Figure 6 is a diagram illustrating the comment display area of ​​an embodiment of the present invention.

[0207] Figure 7 is a diagram illustrating the comment display area of ​​an embodiment of the present invention.

[0208] In one embodiment, the comment display area 500 may be included in the first page 400.

[0209] Comment display area 500 can display more than one comment information 510, 520.

[0210] In one embodiment, the electronic device 101 may apply visual effects to phrases or words corresponding to the selected comment keywords, i.e., keyword-related words 511 and 521, for one or more comment information 510 and 520 displayed in the comment display area 500.

[0211] For example, referring to FIG6, when a user selects a first comment keyword such as "low stimulation" based on the user input corresponding to the first keyword icon 401, comment information corresponding to the first comment keyword can be displayed in the comment display area 500. Furthermore, visual effects such as changing the text font can be applied to the keyword association words 511 and 521 corresponding to the first comment keyword.

[0212] Referring to Figure 7, when the user selects a second review keyword such as "cleaning power" based on the user input corresponding to the second keyword icon 401, visual effects can be applied to the corresponding keyword-related words 512 and 522.

[0213] In one embodiment, comment information 510 and 520 can be displayed in a thumbnail form in the comment display area 500 based on the comment keywords selected by the user based on the keyword icon. For example, the first page information can be generated as a thumbnail comment information by generating a portion of the detailed evaluation information of the area with the highest relevance to the selected comment keyword and displaying it in the comment display area 500.

[0214] The method of the present invention can be implemented by a computer. The computer may include, for example, a processing device. In the present invention, although the steps of the corresponding method are shown and described in a predetermined order, the steps can be performed sequentially or in any combination according to the present invention. In one embodiment, at least some steps may be performed in parallel, repeatedly, or heuristically. The present invention does not preclude variations or modifications to these methods. In one embodiment, at least some steps may be omitted, or other steps may be added.

[0215] Various embodiments of the present invention can be implemented by software recorded on a machine-readable recording medium. The software can be software used to implement various embodiments of the present invention. Programmers in the art to which this invention pertains can deduce the software based on various embodiments of the present invention. For example, the software can be machine-readable commands (e.g., code or code segments) or programs. The device, as an apparatus capable of operating according to commands read from the recording medium, can be, for example, a computer. In one embodiment, the device can be an apparatus of an embodiment of the present invention. In one embodiment, the processor of the device can execute the called command, causing the components of the machine to perform functions corresponding to the corresponding command. In one embodiment, the processor can be a processor of an embodiment of the present invention. The recording medium can refer to all kinds of recording media that can be read by a device. Recording media may include, for example, ROM (Read-Only Memory), RAM (Random Access Memory), CD-ROM (Compact Disk-ROM), magnetic tape, floppy disk, optical data storage devices, etc. In one embodiment, the recording medium may be memory. In one embodiment, the recording medium may also be distributed across computer systems connected via a network. Software may be distributed across computer systems and executed. The recording medium may be a non-transitory recording medium. A non-transitory recording medium means a tangible medium that actually exists regardless of whether the data is stored semi-permanently or temporarily, excluding transiently transmitted signals.

[0216] The technical concept of the present invention has been described above according to various embodiments. However, the technical concept of the present invention includes various substitutions, modifications, and alterations that can be implemented within the scope of common knowledge in the technical field to which the present invention pertains. Furthermore, it should be understood that such substitutions, modifications, and alterations may be included within the scope of the appended patent applications. Embodiments of the present invention can be combined with each other. The embodiments can be combined in various ways depending on the number of cases, and the combined embodiments also fall within the scope of the present invention. [Simplified Explanation of the Diagram]

[0030] Figure 1 illustrates an environment in which an electronic device according to an embodiment of the present invention can be applied. Figure 2 is a block diagram of an electronic device according to an embodiment of the present invention. Figure 3 is a flowchart illustrating a method for displaying comment information of an item according to an embodiment of the present invention. Figure 4 is a diagram illustrating a first page and a comment keyword area according to an embodiment of the present invention. Figure 5 is a flowchart illustrating a method for displaying keyword-related comment information according to an embodiment of the present invention. Figure 6 is a diagram illustrating a comment display area according to an embodiment of the present invention. Figure 7 is a diagram illustrating a comment display area according to an embodiment of the present invention.

Claims

1. A method for displaying review information of an item, implemented via an electronic device, comprising the following steps: obtaining a user's connection request for a first page from a user terminal, the first page including detailed information of a target item; identifying one or more review pieces corresponding to the target item; extracting one or more review keywords based on the one or more review pieces; generating first page information for outputting the first page in such a way that at least a portion of the first page includes a review keyword area for displaying the one or more review keywords; and transmitting the first page information to the user terminal, wherein each of the one or more review pieces corresponding to the target item includes review score information and detailed evaluation information of the target item, and the step of extracting the one or more review keywords includes the following steps: extracting the user's keywords of interest based on the user's purchase history; and extracting the one or more review keywords from the one or more review pieces corresponding to the target item based on the keywords of interest; wherein the step of extracting the user's keywords of interest includes the following steps: Identify the user's purchase history for one or more similar items, where the one or more similar items correspond to the target item among one or more purchased items identified based on the purchase history; identify the user's review history for the one or more similar items purchased by the user; and extract the user's keywords of interest based on the user's review history for the one or more similar items.

2. The method of Request 1, wherein the step of extracting the above-mentioned comment keyword information includes the following steps: selecting one or more candidate comment information, wherein the candidate comment information is at least a portion of one or more comment information corresponding to the above-mentioned target item; and extracting the above-mentioned one or more comment keywords based on the detailed evaluation information of the selected candidate comment information.

3. The method of request item 2, wherein the step of selecting the above candidate comment information includes the following steps: selecting candidate comment information from the above one or more comment information based on the comment score information corresponding to the above one or more comment information.

4. The method of request item 2, wherein the step of selecting the above candidate comment information includes the following steps: selecting candidate comment information from the above one or more comment information based on the length of the detailed evaluation information corresponding to the above one or more comment information.

5. The method of request item 2, wherein the step of selecting the above candidate review information includes the following steps: selecting review information written by one or more customers with similar preferences as candidate review information, the review information being those identified from the above one or more review information based on the information of the above users.

6. The method of Request 1, wherein the step of extracting one or more comment keywords includes the following steps: extracting a plurality of words included in one or more detailed evaluation information corresponding to each of the one or more comment information; calculating the frequency of occurrence corresponding to each of the plurality of words; and extracting comment keywords from the plurality of words based on the frequency of occurrence.

7. The method of claim 6, wherein the step of extracting comment keywords based on the above-mentioned frequency of occurrence includes the following steps: calculating the similarity between the above-mentioned plurality of words; clustering the above-mentioned plurality of words based on the similarity between the above-mentioned plurality of words; extracting one or more candidate keywords based on the above-mentioned clustering results; classifying the above-mentioned plurality of words in a manner corresponding to each of the above-mentioned one or more candidate keywords; calculating the frequency of occurrence of each of the above-mentioned one or more candidate keywords based on the frequency of occurrence of the words corresponding to each of the above-mentioned one or more candidate keywords; and selecting one or more comment keywords from the above-mentioned candidate keywords based on the frequency of occurrence of each of the above-mentioned candidate keywords.

8. The method of claim 6, wherein the step of extracting comment keywords from the plurality of words includes the following steps: setting a weight value corresponding to each of the above-mentioned comment information; extracting comment keywords from the plurality of words based on the above-mentioned frequency of occurrence corresponding to each of the plurality of words and the weight value set in the comment information corresponding to each of the above-mentioned words.

9. The method of request item 8, wherein the weighting values ​​are set based on the writing time of the comment information corresponding to each of the plurality of words.

10. The method of request item 8, wherein the weighting values ​​are set based on the score information of the comment information corresponding to each of the plurality of words.

11. The method of claim 1, wherein the comment keyword area includes keyword icons corresponding to each of the one or more comment keywords, the first page further includes a comment display area that displays comment content in at least a portion of the area corresponding to the comment keyword area, and the step of generating the first page information includes the following steps: selecting a first comment keyword from the one or more comment keywords based on user input corresponding to the keyword icons; identifying keyword-related comment information corresponding to the selected first comment keyword from the one or more comment information corresponding to the target item; and generating the first page information by displaying the keyword-related comment information corresponding to the first comment keyword in the comment display area.

12. The method of claim 11, wherein the step of generating the first page information by displaying the aforementioned keyword-related comment information in the aforementioned comment display area includes the following steps: identifying, in the aforementioned one or more comments, one or more keyword-related comment information including words corresponding to the selected first comment keyword; and selecting a portion of the aforementioned keyword-related comment information based on at least one of the following: writing time, author credibility, and keyword relevance corresponding to each of the aforementioned one or more keyword-related comments; and generating the aforementioned first page information by displaying the selected portion of the aforementioned keyword-related comment information in the aforementioned comment display area.

13. The method of claim 11, wherein the step of generating the first page information by displaying the aforementioned keyword-related comment information in the aforementioned comment display area includes the following steps: identifying a plurality of words included in the aforementioned keyword-related comment information; identifying one or more keyword-related words corresponding to the selected aforementioned comment keyword among the plurality of words; and generating the aforementioned first page information by displaying the keyword-related comment information in the aforementioned comment display area, wherein the keyword-related comment information is obtained by giving visual effects to the identified aforementioned keyword-related words.

14. The method of request 13, wherein the aforementioned visual effects include font changes to the text corresponding to the aforementioned keyword-related words.

15. An electronic device comprising: One or more processors; one or more memory, which store commands to be executed by the one or more processors; and the electronic device is configured such that, when the commands are executed by the one or more processors, the one or more processors perform the method of any one of claims 1 to 14.

16. A non-transitory computer-readable recording medium that records commands, when executed by one or more processors, to cause the one or more processors to perform actions, the commands being configured to cause the one or more processors to perform any one of claims 1 to 14.

Images