KR20210019370A - Apparatus and method for recommending advertisement according to case-based reasoning of context-awareness data - Google Patents

Abstract

The present invention relates to a recording medium and a device and a method for recommending an advertisement in accordance with case learning-based reasoning of context-aware data. According to the present invention, it is possible to provide advertisement recommendation information suitable for a user′s situation based on the result of advertisement recommendation cases regarding a past situation similar to the user′s new situation. The method includes: a step of collecting and analyzing user-related situation information by means of a context-aware data analysis module; a step of measuring the similarity between the situation information and advertisement recommendation cases stored in a database by means of a similarity measuring module; a step of selecting an advertisement recommendation case related to the user′s situation based on the similarity by means of a selection module; a step of distinguishing a part of the selected advertisement recommendation case requiring correction and correcting the part based on the situation information by means of a correction module to generate an advertisement search word; and a step of recommending an advertisement suitable for the user′s situation in accordance with the advertisement search word by means of an advertisement recommendation module.

Description

Advertisement recommendation device and method according to case-learning-based reasoning of context-aware data {APPARATUS AND METHOD FOR RECOMMENDING ADVERTISEMENT ACCORDING TO CASE-BASED REASONING OF CONTEXT-AWARENESS DATA}

The present invention relates to an advertisement recommendation apparatus and method based on case-learning-based reasoning of context-aware data, and more specifically, a situation in which an advertisement suitable for a user's situation is provided based on past advertisement recommendation cases similar to the user's situation. It relates to an advertisement recommendation apparatus and method according to case learning-based inference of cognitive data.

In general, advertisements are used to widely inform consumers of information on products or services through various media. Appropriate advertisements can contribute to increased sales of sellers or service providers, and can provide an opportunity for consumers to purchase products or services they need.

In recent years, the advertising market has reached an oversupply state due to the ease of access to online shopping malls, blogs, and social network services (SNS). The consumer wants to selectively receive advertisements for the goods or services he needs, but as a result of indiscriminately exposing various advertisements outside of interest to the consumer, it may become difficult for the consumer to purchase appropriate products or provide services. .

From the advertiser's point of view, when advertisements are exposed indiscriminately to users who do not have demand for the product or service as well as consumers who are highly likely to purchase the product or use the service, the cost of advertising is increased due to unnecessary advertisement exposure. There is.

In addition, most of the existing advertisements are executed as many times as desired by the advertiser at a time desired by the advertiser without consideration of the situation or environment around the consumer. Therefore, prior research is required before uploading an advertisement, and there is a problem in that advertisement costs increase due to such preliminary research costs.

The present invention is an advertisement recommendation apparatus and method according to case learning-based inference of situation awareness data for selecting and recommending advertisements suitable for the user's situation, based on past advertisement recommendation cases similar to the user's situation such as weather, traffic, and time, It is to provide a recording medium.

In addition, the present invention compares the user's situation with the past advertisement recommendation cases, corrects some of the advertisement recommendation cases, and provides an advertisement based on the modified case, so that the situation can recommend an advertisement more suitable for the user's situation. It is to provide an advertisement recommendation apparatus, method, and recording medium based on case learning-based inference of data.

In addition, the present invention matches and evaluates appropriate advertisements according to the context information collected from the user, learns recommended advertisements according to the user's context, and accumulates context-specific cases. It is to provide an advertisement recommendation apparatus and method, and a recording medium based on case learning-based inference of context-aware data that can recommend a situation.

An advertisement recommendation apparatus according to case learning-based inference of context-aware data according to an embodiment of the present invention includes: a database storing advertisement recommendation cases for various user types and various situations; A context-aware data analysis module that collects and analyzes various context information related to users; A similarity measurement module for measuring a similarity between the various context information analyzed by the context-aware data analysis module and advertisement recommendation cases stored in the database; A selection module for selecting an advertisement recommendation case related to the user's situation based on the similarity measured by the similarity measurement module; A correction module for generating an advertisement search word by determining a portion to be corrected among the ad recommendation cases selected by the selection module, and correcting the portion to be corrected based on the various context information; And an advertisement recommendation module for recommending an advertisement suitable for the user's situation according to the advertisement search word modified by the modification module.

An advertisement recommendation device based on case learning-based inference of context-aware data according to an embodiment of the present invention includes: a verification module for verifying whether a recommended advertisement provided by the advertisement search word is suitable for the user's situation; And a learning module for updating the advertisement recommendation cases in the database by learning based on a verification result of the advertisement search word.

The verification module may verify the recommended advertisement according to the situation of the user based on sales information generated according to the recommended advertisement. The learning module may update the advertisement recommendation cases according to the user's feedback data on the recommendation advertisement.

The context-aware data analysis module may collect context information including geographic information, weather information, traffic information, time information, personal information of the user, and a search word input by the user to receive advertisement recommendations. have.

The correction module determines a portion inconsistency between the advertisement recommendation case and the various context information as the correction required portion; The advertisement search word may be generated by modifying the portion to be corrected based on context information corresponding to the inconsistent portion.

An advertisement recommendation method according to case-learning-based reasoning of context-aware data according to an embodiment of the present invention includes: collecting and analyzing various context information related to a user by a context-aware data analysis module; Measuring a similarity between the various context information and advertisement recommendation cases stored in a database, by a similarity measurement module; Selecting, by a selection module, an advertisement recommendation case related to the user's situation based on the similarity; Generating an advertisement search word by determining, by a correction module, a portion to be corrected among the ad recommendation cases, and modifying the portion to be corrected based on the various context information; And recommending, by the advertisement recommendation module, an advertisement suitable for the situation of the user according to the advertisement search word.

An advertisement recommendation method according to case learning-based reasoning of context-aware data according to an embodiment of the present invention includes, by a verification module, verifying whether a recommended advertisement recommended by the advertisement search word is suitable for the user's situation; And updating, by the learning module, the advertisement recommendation cases in the database by learning based on the verification result of the advertisement search word.

The verifying may include verifying a recommended advertisement according to a situation of the user based on sales information generated according to the recommended advertisement. The step of updating the advertisement recommendation cases may include updating the advertisement recommendation cases according to the user's feedback data on the recommendation advertisement.

The collecting and analyzing of the various situation information may include geographical information, weather information, traffic information, time information, personal information of the user, and a search word input by the user to receive advertisement recommendation. It may include collecting them.

The generating of the advertisement search word may include determining a portion inconsistency between the advertisement recommendation case and the various context information as the correction required portion; And generating the advertisement search word by modifying the portion that needs to be corrected based on context information corresponding to the inconsistent portion.

According to an embodiment of the present invention, there is provided a computer-readable recording medium in which a program for executing an advertisement recommendation method according to case learning-based reasoning of the context-aware data is recorded.

According to an embodiment of the present invention, based on past advertisement recommendation cases similar to the user's situation such as weather, traffic, time, etc., advertisements based on case learning-based inference of context-aware data for selecting and recommending advertisements suitable for the user's situation A recommended apparatus and method, and a recording medium are provided.

In addition, according to an embodiment of the present invention, by comparing a user's situation and past advertisement recommendation cases, modifying some of the advertisement recommendation cases, and providing advertisements based on the modified cases, recommending advertisements more suitable for the user's situation. An advertisement recommendation apparatus and method, and a recording medium according to case-learning-based reasoning of possible situational data are provided.

In addition, according to an embodiment of the present invention, by matching and evaluating appropriate advertisements according to the context information collected from the user, learning recommended advertisements according to the user's context, and accumulating case-specific cases, the more the user is You can recommend advertisements suitable for your situation.

1 is a block diagram of an advertisement recommendation apparatus based on case learning-based inference of context-aware data according to an embodiment of the present invention.
2 is a configuration diagram of a context-aware data analysis module constituting an advertisement recommendation device based on case learning-based inference of context-aware data according to an embodiment of the present invention.
3 is a flowchart of an advertisement recommendation method according to case learning-based inference of context-aware data according to an embodiment of the present invention.
4 is a flowchart of step S150 of FIG. 3.

Advantages and features of the present invention, and a method of achieving them will become apparent with reference to the embodiments described below in detail together with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but may be implemented in a variety of different forms, and only these embodiments make the disclosure of the present invention complete, and are common knowledge in the technical field to which the present invention pertains. It is provided to completely inform the scope of the invention to those who have, and the invention is only defined by the scope of the claims. The same reference numerals refer to the same components throughout the specification.

In the present specification, when a part "includes" a certain component, it means that other components may be further included rather than excluding other components unless otherwise stated. As used herein,'~ unit' is a unit that processes at least one function or operation, and may mean, for example, software, FPGA, or hardware components. The functions provided by the'~ unit' may be performed separately by a plurality of elements, or may be integrated with other additional elements. The'~ unit' in the present specification is not necessarily limited to software or hardware, and may be configured to be in an addressable storage medium, or may be configured to reproduce one or more processors.

The apparatus and method for recommending advertisements according to case-learning-based reasoning of context-aware data according to an embodiment of the present invention is based on the results of advertisement recommendation cases for a past situation similar to a new situation of a user, and case-learning-based inference of context-aware data. According to the user's situation, it is possible to recommend an advertisement suitable for the situation.

The apparatus and method for recommending advertisements according to case-learning-based inference of context-aware data according to an embodiment of the present invention first collects and analyzes various context information related to users (eg, weather, traffic, time, etc.), and After measuring the similarity between various context information of and the advertisement recommendation cases stored in the database, one or more advertisement recommendation cases related to the user's situation may be selected based on the similarity.

Next, among the selected ad recommendation cases, the part that needs to be corrected that is inconsistent with the user's situation is determined, the part that needs to be modified is modified based on the user's situation information to generate an ad search word, and the user's situation according to the modified ad search term You can recommend an advertisement suitable for

Afterwards, it is possible to verify whether the recommended advertisement recommended by the advertisement search term is suitable for the user's situation through sales comparison, etc., and the weight of the case can be adjusted by learning based on the verification result of the advertisement search term, and new advertisement recommendation The cases can be updated in the database.

According to an embodiment of the present invention, an advertisement suitable for a user's situation may be selected and recommended based on results of advertisement recommendation cases related to a past situation similar to the user's situation such as weather, traffic, and time.

In addition, according to an embodiment of the present invention, by comparing a user's situation and past advertisement recommendation cases, modifying some of the advertisement recommendation cases, and providing advertisements based on the modified cases, recommending advertisements more suitable for the user's situation. can do.

In addition, according to an embodiment of the present invention, by matching and evaluating suitable advertisements according to the collected situation information, learning recommended advertisements according to the situation, and accumulating cases for each situation, so that as data is accumulated, advertisements suitable for similar situations are automatically generated. I can recommend it.

1 is a block diagram of an advertisement recommendation apparatus based on case learning-based inference of context-aware data according to an embodiment of the present invention. Referring to FIG. 1, an advertisement recommendation device 100 according to case learning-based inference of context-aware data according to an embodiment of the present invention includes a database 110, a context-aware data analysis module 120, and a similarity measurement module 130. , A selection module 140, a modification module 150, an advertisement recommendation module 160, a verification module 170, and a learning module 180.

The database 110 may store advertisement recommendation cases related to various user types and various situations. The database 110 may store caseized data previously input by the administrator for context analysis and advertisement recommendation.

The database 110 is a volatile memory such as SRAM, DRAM, SDRAM, ROM, PROM, EPROM, EEPROM, flash memory, PRAM, MRAM, RRAM, nonvolatile memory such as FRAM, a hard disk or an optical reading medium, for example, a CDROM. , An optical recording medium such as a DVD, and a storage medium such as a magnetic medium such as a magnetic tape, but are not limited thereto.

2 is a configuration diagram of a context-aware data analysis module constituting an advertisement recommendation device based on case learning-based inference of context-aware data according to an embodiment of the present invention. 1 and 2, the context-aware data analysis module 120 may collect and analyze various context information related to a user.

The context-aware data analysis module 120 stores context information including geographic information, weather information, traffic information, time information, user's personal information (or user type), and a search word input by the user to receive advertisement recommendations. Can be collected. The context of the user may be analyzed based on various context data collected by the context-aware data analysis module 120.

In an embodiment, the context-aware data analysis module 120 includes a location information collection unit 121, a weather information collection unit 122, a traffic information collection unit 123, a time information collection unit 124, and a personal information collection unit ( 125), may include a search word collection unit 126, and the like.

The location information collection unit 121 may collect current location information (geographic information) of a user. The weather information collection unit 122 may collect weather information on the user's current location. The traffic information collection unit 123 may collect traffic information related to the user's current location.

The time information collection unit 124 may collect time information related to a user's situation. The personal information collection unit 125 may collect personal information of the user (eg, gender, user type such as age, birthday, etc.). The search word collection unit 126 may collect a search word input by a user.

In an embodiment, the user's context information may be collected by the user terminal. The user terminal may be implemented as, for example, a smart phone, a tablet PC, a notebook, a desktop PC, a laptop PC, a netbook computer, a workstation, a PDA, a PMP, a smart glasses, or a wearable device such as a smart band, but is limited thereto. no.

The user's current location may be collected by, for example, a Global Positioning System (GPS) built into the user terminal, or may be collected by location information input by the user into the user terminal.

The user's personal information may be collected from various pieces of information (eg, age, gender, birthday, etc.) set by the user in an advertisement recommendation app executed on the user terminal.

When the user's current location is determined, the context-aware data analysis module 120 may collect weather information, traffic information, and the like for the user's current location through a network search.

Time information related to the user's situation may be collected from the current time or a search word input by the user to the user terminal. In addition, the context-aware data analysis module 120 may collect a search word input by a user to receive advertisement recommendation through a user terminal.

The similarity measurement module 130 may measure a similarity between various context information of a user collected by the context-aware data analysis module 120 and advertisement recommendation cases stored in the database 110, respectively.

The similarity between user context information and pre-stored ad recommendations is based on, for example, Euclidean Distance, Manhattan Distance, Minkowski Distance, and Cosine Similarity. It may be calculated as, but is not limited thereto.

For example, user context information and pre-stored advertisement recommendation cases may be vectorized in a preset order for each type of context information, and a similarity between two vectors may be calculated, but is not limited to this method. For example, various changes, such as selecting an advertisement recommendation case, can be made by searching the database based on the search word related to the user's context information.

The selection module 140 may select an advertisement recommendation case related to a user's situation based on the similarity measured by the similarity measurement module 130. The selection module 140 may select one or a plurality of advertisement recommendation cases as a case similar to the current situation of the user in the order of high similarity to the user's situation.

When a target case (user's situation) is given, the selection module 140 may search/select an appropriate advertisement recommendation case to provide advertisement recommendation information for the target case in the database 110 in which the target case is stored.

In this way, based on the similarity measure indicating the degree of similarity between the user's situation and past advertisement recommendation cases, an appropriate advertisement recommendation case related to the user's situation may be searched in the database 110.

In this case, the case stored in the database 110 may include context information, a recommendation advertisement for the situation, and a description of how the recommendation advertisement was induced.

The modification module 150 may generate an advertisement search word by determining a part that needs to be modified among the ad recommendation cases selected by the selection module 140, and corrects the part that needs to be modified based on various context information related to a user.

The correction module 150 extracts, for example, a keyword that does not correspond to the user's situation or has a low relevance among keywords included in the ad recommendation case, as a part that needs to be modified, and then selects a keyword corresponding to the part that needs to be modified. Advertising search words (modified keywords) can be generated by modifying to suit the situation.

The modification module 150 determines whether the ad recommendation case selected by the selection module 140 can be applied immediately, and if it is difficult to apply immediately, the modification module 150 may generate a modified case (modified advertisement recommendation case) by modifying past advertisement recommendation cases. I can.

If the case selected by the selection module 140 may be applied immediately, the advertisement recommendation case is reused, and the solution of the recommendation advertisement problem for the target situation is mapped from the previous advertisement recommendation case. .

To this end, the modification module 150 may include adapting a solution as necessary to fit a new situation. The modification module 150 may connect the solution to the situation to the target situation based on the previous case, and then modify the advertisement recommendation case if necessary.

Meanwhile, the correction module 150 may be provided to directly correct a case of data that does not fit the user's situation among the recommended data by input through the manager's manager terminal.

In an embodiment, the correction module 150 determines the inconsistent portion between the advertisement recommendation case and various context information related to the user as a portion to be corrected, and corrects the determined portion to be corrected based on context information corresponding to the inconsistent portion By doing so, you can create an advertisement search word.

The advertisement recommendation module 160 may provide a recommended advertisement through a user terminal or an advertisement device around the user by matching advertisements suitable for various context information of a user according to an advertisement search word determined by the correction module 150.

The verification module 170 may verify whether the recommended advertisement recommended by the advertisement search word is suitable for the user's situation. The verification module 170 may test the result of the recommended advertisement provided according to the user's situation in the real world or through a simulation.

The verification module 170 may perform verification by comparing the application result of the previous case and the application result of the current case. The verification module 170 may perform case verification by determining whether it is worth preserving the application result of this case as a new advertisement recommendation case through, for example, sales comparison.

The learning module 180 may update the ad recommendation cases in the database 110 by setting a weight of a corresponding case by learning based on a verification result of an advertisement search word (S180). The learning module 180 may update advertisement recommendation cases according to user feedback data for the recommendation advertisement.

The user's feedback data may be collected through a positive/negative reaction to the recommended advertisement input through the user terminal (eg, clicking a "like" icon, inputting reviews, etc.).

After verification by the verification module 170, the learning module 180 may evaluate an advertisement recommendation result according to a case, and then adjust the case-based index and weight as necessary to reflect the advertisement recommendation.

When it is determined that the recommendation advertisement for the user's situation is successfully applied, a new advertisement recommendation experience may be stored in the database 110 as a new advertisement recommendation case.

According to an embodiment of the present invention, unlike an existing advertisement system that recommends advertisements only according to predetermined conditions, an advertisement recommendation case for a past situation similar to the user's situation by analyzing various situations such as weather, traffic, and time of the user Based on the results, it is possible to select and recommend an advertisement suitable for the user's corresponding situation.

Therefore, by providing appropriate advertisements according to the user's situation, improved advertisement effects can be expected, and advertisements can be prevented from being provided to users indiscriminately, allowing users to efficiently make decisions on product purchase or service use. .

In addition, according to an embodiment of the present invention, by matching and evaluating appropriate advertisements according to the collected context information, learning appropriate recommended advertisements according to the user's situation, accumulating cases for each context, and providing more suitable advertisements in a similar environment in the future. Matching and recommending.

The advertisement recommendation result may be evaluated by the manager (or advertiser) and reflected in the system. As the process is repeated, the manager (or advertiser)'s intention can be reflected in the advertisement recommendation through the manager's intention learning.

2 is a flowchart of an advertisement recommendation method based on case learning-based inference of context-aware data according to an embodiment of the present invention. 1 and 2, the database 110 may store advertisement recommendation cases for various user types and various situations (S110). The database 110 may store instanced data previously input by an administrator for analysis.

The context-aware data analysis module 120 may collect and analyze various context information related to a user (S120). The context-aware data analysis module 120 may collect context information, including geographic information, weather information, traffic information, time information, personal information of the user, and a search word input by the user to receive advertisement recommendation.

Next, the similarity measurement module 130 may measure the similarity between various context information of the user collected/analyzed by the context-aware data analysis module 120 and advertisement recommendation cases stored in the database 110 (S130). ).

The similarity between user context information and ad recommendations is calculated based on, for example, Euclidean Distance, Manhattan Distance, Minkowski Distance, and Cosine Similarity. It may be, but is not limited thereto.

Next, the selection module 140 may select an advertisement recommendation case related to the user's situation based on the similarity measured by the similarity measurement module 130 (S140). The selection module 140 may select one or a plurality of advertisement recommendation cases as a case similar to the current situation in order of high similarity to the user's situation.

When a target case (situation) is given, the selection module 140 may search for an appropriate advertisement recommendation case to provide advertisement recommendation information for the target case in a database in which the case is stored. An advertisement recommendation case may consist of context information, a recommendation advertisement for the advertisement, and a description of how the recommendation advertisement was induced.

The correction module 150 may generate an advertisement search word by determining a portion that needs to be corrected among the selected advertisement recommendation cases, and corrects the portion that needs to be corrected based on various context information related to a user (S150).

The correction module 150 determines whether the selected case may be applied immediately, and if it is difficult to apply immediately, the modification module 150 may modify the case to generate a modified case. If the selected case can be applied immediately, the case is reused, and the solution of the recommendation advertisement problem for the target situation is linked from the previous case.

The modification module 150 may include adapting the solution as needed to fit the new situation. The modification module 150 may connect a solution to the situation to the target situation based on the previous case, and may modify it if necessary.

4 is a flowchart of step S150 of FIG. 3. Referring to FIGS. 1, 3 and 4, the correction module 150 may first determine a portion inconsistent between an advertisement recommendation case and various context information related to a user as a portion requiring correction (S152 ). The correction module 150 may generate an advertisement search word by correcting the determined portion requiring correction based on context information corresponding to the inconsistent portion (S154).

Referring back to FIGS. 1 and 2, the advertisement recommendation module 160 may recommend an advertisement suitable for a user's situation according to an advertisement search word (S160). The verification module 170 may test a new advertisement recommendation case in the real world or by simulation, and verify whether the recommended advertisement recommended by the advertisement search word is suitable for the user's situation (S170).

The verification module 170 may perform verification by comparing the application result of the previous case and the application result of the current case. The verification module 170 may compare whether it is worth preserving the application result of this case as a new case, for example, by comparing sales according to the application result of the previous case and the current case.

The learning module 180 may update ad recommendation cases in the database 110 by learning based on the verification result of the advertisement search word (S180). The learning module 180 may update advertisement recommendation cases according to user feedback data for the recommendation advertisement.

After verification by the verification module 170, the learning module 180 evaluates the advertisement recommendation result according to the case, and then adjusts the index and weight related to various situations of the advertisement recommendation case as necessary to reflect it in the advertisement recommendation. have. When a recommendation advertisement for a user's situation is successfully applied, a new advertisement recommendation experience may be stored in the database 110 as a new advertisement recommendation case.

Hereinafter, an example of using a service according to an embodiment of the present invention will be described. As the first example, the process of recommending advertisements in the case of "a rainy evening, when a young man comes alone" is described as an example. Search cases according to the keywords "rainy", "dinner", "young man", and "alone" to recommend the menu and set advertisements (eg, electronic menu board, large screen, etc.).

At this time, first, the previously used "more delicious wine list on a rainy day", "a good drink recommended for eating alone at the store", and "good food recommendations to eat when the body is cold" can be called. Then, based on the keywords "alone" and "young man" corresponding to the user's situation, the "recommendation of liquor that is good to eat alone at the store" is modified to "recommendation of liquor that is good for appealing to the opposite sex in the store", and the menu based on this You can recommend and send out advertisements.

In addition, by comparing the sales of the previous case and the new case, it is compared with the previous keyword to see if this keyword is more suitable for a man who came alone, and if it is determined that the keyword is suitable, the case is saved as an additional case. , Compared with the previous case, the weights can be adjusted.

As a second example, an example of recommending an advertisement in the context of "Friends gathered in the evening to celebrate a friend's birthday" is described. First, you can search with keywords such as "birthday", "dinner", and "friends" to recommend menus suitable for the keywords "birthday", "dinner", and "friends" and set advertisements.

At this time, the ad recommendation cases such as "Happy Birthday Song displayed on the large screen and a party atmosphere to enjoy together", "Good food to eat alone in the evening", "Recommended food that can be shared by multiple people" were called, and then, Based on keywords such as "dinner" and "friends" corresponding to the user's situation, the ad recommendation case was modified from "good food for dinner alone" to "food recommended for dinner party atmosphere", and the menu based on this You can recommend and send out advertisements.

In addition, by comparing the sales in the previous case with the sales in the new case, when comparing with the previous keyword, it is checked whether this keyword is more suitable for a man who has come alone, and the case is saved as an additional case, and You can compare and adjust the weights.

According to an embodiment of the present invention, by matching and evaluating appropriate advertisements according to the collected context data case, and learning recommended advertisements according to the context, as the data accumulates, it is possible to recommend more suitable advertisements for the context.

According to an embodiment of the present invention, advertisements can be selected and recommended according to collectible conditions (situation information) such as weather, traffic, and time. In addition, it is possible to automatically recommend advertisements in a similar environment later by accumulating case-specific cases based on the recommended data.

According to an embodiment of the present invention, based on the results of advertisement recommendation cases for a past situation similar to a new situation of a user, advertisement recommendation information suitable for a corresponding situation of the user may be provided.

According to an embodiment of the present invention, unlike an existing advertisement system that recommends advertisements only according to predetermined conditions, it is possible to recommend appropriate advertisements according to the case by analyzing a user's situation.

The advertisement recommendation result can be evaluated by the manager and reflected in the system. As the process is repeated, the manager's intention can be reflected in the advertisement recommendation through the learning of the manager's intention.

According to an embodiment of the present invention, rather than an advertisement target analysis and advertisement progress method, which was an existing business model, various attributes of advertisement can be set to recommend advertisements suitable for a situation, and improved advertisement effects can be expected through flexible response to situations.

The embodiments described above may be implemented as a hardware component, a software component, and/or a combination of a hardware component and a software component. For example, the devices, methods, and components described in the embodiments are, for example, a processor, a controller, an Arithmetic Logic Unit (ALU), a digital signal processor, a microcomputer, a field programmable gate (FPGA). Array), Programmable Logic Unit (PLU), microprocessor, or any other device capable of executing and responding to instructions, such as one or more general purpose computers or special purpose computers.

The processing device may execute an operating system and one or more software applications executed on the operating system. In addition, the processing device may access, store, manipulate, process, and generate data in response to the execution of software. For the convenience of understanding, although it is sometimes described that one processing device is used, those of ordinary skill in the art may use a processing device that includes a plurality of processing elements and/or multiple types of processing elements. It will be understood that it can include.

For example, the processing device may include a plurality of processors or one processor and one controller. In addition, other Processing configurations are possible, such as a Parallel Processor. The software may include a computer program, code, instruction, or a combination of one or more of them, and configure the processing unit to operate as desired, or process it independently or collectively. You can command the device.

Software and/or data may be interpreted by a processing device or to provide instructions or data to a processing device, of any type of machine, component, physical device, virtual equipment, computer storage medium or device. , Or may be permanently or temporarily embodyed in a transmitted signal wave. The software may be distributed over networked computer systems and stored or executed in a distributed manner. Software and data may be stored on one or more computer-readable recording media.

The method according to the embodiment may be implemented in the form of program instructions that can be executed through various computer means and recorded in a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, and the like alone or in combination. The program instructions recorded on the medium may be specially designed and configured for the embodiment, or may be known and usable to those skilled in computer software.

Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks and magnetic tapes, optical media such as CDROMs and DVDs, and ROM, RAM, and flash memory. Hardware devices specially configured to store and execute program instructions such as, etc. are included. Examples of the program instructions include not only machine language codes such as those produced by a compiler, but also high-level language codes that can be executed by a computer using an interpreter or the like. The hardware device described above may be configured to operate as one or more software modules to perform the operation of the embodiment, and vice versa.

Although the embodiments have been described by the limited embodiments and drawings as described above, various modifications and variations can be made from the above description to those of ordinary skill in the art. For example, the described techniques are performed in a different order from the described method, and/or components such as a system, structure, device, circuit, etc. described are combined or combined in a form different from the described method, or other components Alternatively, even if substituted or substituted by an equivalent, an appropriate result can be achieved. Therefore, other implementations, other embodiments, and equivalents to the claims fall within the scope of the following claims.

100: Ad recommendation device based on case-learning-based reasoning of context-aware data
110: database
120: context-aware data analysis module
121: location information collection unit
122: Weather information collection unit
123: Traffic information collection department
124: time information collection unit
125: personal information collection unit
126: search term collection unit
130: similarity measurement module
140: selection module
150: modification module
160: ad recommendation module
170: verification module
180: learning module

Claims (11)

A database storing advertisement recommendation cases for various user types and various situations;
A context-aware data analysis module that collects and analyzes various context information related to users;
A similarity measurement module for measuring a similarity between the various context information analyzed by the context-aware data analysis module and advertisement recommendation cases stored in the database;
A selection module for selecting an advertisement recommendation case related to the user's situation based on the similarity measured by the similarity measurement module;
A correction module for generating an advertisement search word by determining a portion to be corrected among the ad recommendation cases selected by the selection module, and correcting the portion to be corrected based on the various context information; And
Comprising an advertisement recommendation module for recommending an advertisement suitable for the user's situation according to the advertisement search word modified by the modification module,
Ad recommendation device based on case-learning-based reasoning of context-aware data. The method of claim 1,
A verification module for verifying whether the recommended advertisement provided by the advertisement search word is suitable for the user's situation; And
Further comprising a learning module for updating the advertisement recommendation cases in the database by learning based on the verification result of the advertisement search word,
Ad recommendation device based on case-learning-based reasoning of context-aware data. The method of claim 2,
The verification module verifies the recommended advertisement according to the user's situation based on sales information generated according to the recommendation advertisement,
The learning module updates the advertisement recommendation cases according to the user's feedback data on the recommendation advertisement,
Ad recommendation device based on case-learning-based reasoning of context-aware data. The method according to any one of claims 1 to 3,
The context-aware data analysis module collects context information, including geographic information, weather information, traffic information, time information, personal information of the user, and a search word input by the user to receive advertisement recommendation related to the user,
Ad recommendation device based on case-learning-based reasoning of context-aware data. The method according to any one of claims 1 to 3,
The modification module,
Discriminating a portion inconsistency between the advertisement recommendation case and the various context information as the correction required portion;
Generating the advertisement search word by modifying the portion requiring correction based on context information corresponding to the inconsistent portion,
Ad recommendation device based on case-learning-based reasoning of context-aware data. Collecting and analyzing various context information related to a user by a context-aware data analysis module;
Measuring a similarity between the various context information and advertisement recommendation cases stored in a database, by a similarity measurement module;
Selecting, by a selection module, an advertisement recommendation case related to the user's situation based on the similarity;
Generating an advertisement search word by determining, by a correction module, a portion to be corrected among the ad recommendation cases, and modifying the portion to be corrected based on the various context information; And
Including, by the advertisement recommendation module, recommending an advertisement suitable for the situation of the user according to the advertisement search word,
Ad recommendation method based on case-learning-based reasoning of context-aware data. The method of claim 6,
Verifying, by a verification module, whether the recommended advertisement recommended by the advertisement search word is suitable for the user's situation; And
Further comprising, by a learning module, updating the advertisement recommendation cases in the database by learning based on the verification result of the advertisement search word,
Ad recommendation method based on case-learning-based reasoning of context-aware data. The method of claim 7,
The verifying step includes verifying a recommended advertisement according to the user's situation based on sales information generated according to the recommended advertisement,
The step of updating the advertisement recommendation cases includes updating the advertisement recommendation cases according to the user's feedback data on the recommended advertisement,
Ad recommendation method based on case-learning-based reasoning of context-aware data. The method of claim 6,
The collecting and analyzing of the various situation information may include geographical information, weather information, traffic information, time information, personal information of the user, and a search word input by the user to receive advertisement recommendation. Comprising collecting them,
Ad recommendation method based on case-learning-based reasoning of context-aware data. The method of claim 6,
Generating the advertisement search word,
Determining a part inconsistency between the advertisement recommendation case and the various context information as the part requiring correction; And
Comprising the step of generating the advertisement search word by modifying the part that needs to be modified based on context information corresponding to the inconsistent part,
Ad recommendation method based on case-learning-based reasoning of context-aware data. A computer-readable recording medium in which a program for executing an advertisement recommendation method based on case learning-based reasoning of the situational awareness data according to any one of claims 6 to 10 is recorded.

Images