KR20220055400A - Method for searching a real estate area and property based on personal residence valuation criteria using artificial intelligence - Google Patents

Abstract

By using a preferred area description and preferred sale property description of a user and searching for a real estate area and sale property that meet a personal residence valuation evaluation criteria of the user, a method for searching the real estate area and sale property based on the personal residence valuation evaluation criteria using an artificial intelligence according to a preferred embodiment of the present invention allows a customizable real estate area and sale property for each user to be searched/recommended, thereby improving a satisfaction of the user.

Description

{Method for searching a real estate area and property based on personal residence valuation criteria using artificial intelligence}

The present invention relates to a real estate area and property search method based on personal housing value evaluation criteria using artificial intelligence, and more particularly, to search for a real estate area and property, based on personal housing value evaluation criteria using artificial intelligence, real estate area and property search it's about how

Conventional real estate for sale search is based on transaction type, floor area, price, and other photos. Recently, a method of searching for properties based on a typical residential value concept in consideration of the commute distance to work using GIS (Geographic Information System) information, a station area, a commercial area, a school district, etc. has been provided.

However, the housing market is changing from supplier-centric to consumer-centric, and the criteria by which people evaluate housing value are very specific and subdivided in the traditional housing type, size, price, station area, green area, etc. Factors that previously did not give much weight as residential value, such as accessibility to small concert halls and rooftop presence, are considered as important factors in choosing a residential area or real estate, so it is subdivided into the conventional real estate sale database information and individual housing value There is a problem in that it is difficult to search for preferred residential areas and real estate properties that meet the evaluation criteria.

An object of the present invention is to search for a real estate area and property that meet the user's individual housing value evaluation criteria using the user's preferred area description and preferred sale description based on artificial intelligence, personal housing value evaluation using artificial intelligence It is to provide a criteria-based real estate area and property search method.

Other objects not specified in the present invention may be additionally considered within the scope that can be easily inferred from the following detailed description and effects thereof.

The real estate area and property search method based on individual housing value evaluation criteria using artificial intelligence according to a preferred embodiment of the present invention for achieving the above object may not be considered in the general housing value evaluation criteria using artificial intelligence, In order to search real estate areas and properties based on the housing value evaluation that is important to an individual, area-related unstructured natural language information is collected through a web crawling technique from open sources of information, and the collected unstructured natural language information constructing a local information corpus DB by processing; receiving a preferred area description and a preferred sale description for the real estate area and sale through the user terminal; Based on the regional information corpus DB, based on the preferred region description, semantic similarity is analyzed through an artificial intelligence natural language processing technique, and preference is given along with regional similarity in the order of the preferred region description and the region with the highest regional similarity. providing a local candidate search result to the user terminal; receiving a preferred area selected from the preferred area candidate search result through the user terminal; And based on the real estate for sale DB in which information on real estate for sale is stored, in the selected preferred area, by analyzing the similarity of the real estate sale description and real estate sale photo through artificial intelligence technique, the preferred sale description and the sale with high similarity order and providing, to the user terminal, a preferred listing candidate search result together with the similarity of the listing.

Here, the step of constructing the regional information corpus DB includes analyzing the collected atypical natural language information for each atypical natural language information through an artificial intelligence natural language processing technique, and including a regional item, an alias item, an information input year item, and a regional description item This may be accomplished by building the local information corpus DB.

Here, the step of providing the preferred region candidate search result includes, for each region of the region information corpus DB, the semantic similarity between the information content item of the region description item and the preferred region description, and the sentiment analysis item of the region description item. As a result of the sentiment evaluation, the regional similarity is obtained by region by acquiring the regional similarity based on the correction constant according to the information depreciation and the information source reliability of the region description item obtained based on the information input year item and the search year. can be done

Here, the preferred sale candidate search result providing step includes searching the real estate for sale corresponding to the selected preferred area in the real estate for sale DB, and for each searched real estate for sale, based on the real estate sale basic information and the preferred sale description Based on the obtained basic information similarity, the real estate sale description and the description similarity obtained based on the preferred sale description, and the real estate sale photo and drawing and the photographic similarity obtained based on the preferred sale description, the searched real estate sale similarity level can be achieved by obtaining

According to the real estate area and property search method based on personal residential value evaluation criteria using artificial intelligence according to a preferred embodiment of the present invention, a real estate area that meets the user's individual housing value evaluation criteria using the user's preferred area description and preferred property description And by searching for sale, it is possible to search/recommend customized real estate areas and properties for each user, thereby improving user satisfaction.

Effects of the present invention are not limited to the above-mentioned effects, and other effects not mentioned will be clearly understood by those skilled in the art from the following description.

1 is a block diagram for explaining a real estate area and property search apparatus based on personal residential value evaluation criteria using artificial intelligence according to a preferred embodiment of the present invention.
2 is a flowchart illustrating a method for searching a real estate area and for sale based on a personal residence value evaluation criterion using artificial intelligence according to a preferred embodiment of the present invention.
3 is a view for explaining an example of a real estate area and property search process based on personal residential value evaluation criteria using artificial intelligence according to a preferred embodiment of the present invention.
4 is a diagram illustrating an example of a real estate area and property search process based on personal residential value evaluation criteria using artificial intelligence according to a preferred embodiment of the present invention.
FIG. 5 is a diagram illustrating in more detail the preferred area search process shown in FIG. 4 .
6 is a view showing in more detail the process of searching for a preferred item shown in FIG. 4 .

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. Advantages and features of the present invention, and a method for achieving them will become apparent with reference to the embodiments described below in detail in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments published below, but may be implemented in various different forms, and only these embodiments make the publication of the present invention complete, and common knowledge in the art to which the present invention pertains It is provided to fully inform those who have the scope of the invention, and the present invention is only defined by the scope of the claims. Like reference numerals refer to like elements throughout.

Unless otherwise defined, all terms (including technical and scientific terms) used herein may be used with the meaning commonly understood by those of ordinary skill in the art to which the present invention belongs. In addition, terms defined in a commonly used dictionary are not to be interpreted ideally or excessively unless clearly defined in particular.

In the present specification, terms such as “first” and “second” are for distinguishing one component from other components, and the scope of rights should not be limited by these terms. For example, a first component may be termed a second component, and similarly, a second component may also be termed a first component.

In the present specification, identification symbols (eg, a, b, c, etc.) in each step are used for convenience of description, and identification symbols do not describe the order of each step, and each step is clearly Unless a specific order is specified, the order may differ from the specified order. That is, each step may occur in the same order as specified, may be performed substantially simultaneously, or may be performed in the reverse order.

In this specification, expressions such as “have”, “may have”, “include” or “may include” indicate the existence of a corresponding feature (eg, a numerical value, function, operation, or component such as a part). and does not exclude the presence of additional features.

In addition, the term '~ unit' as used herein means software or a hardware component such as a field-programmable gate array (FPGA) or ASIC, and '~ unit' performs certain roles. However, '-part' is not limited to software or hardware. '~' may be configured to reside on an addressable storage medium or may be configured to refresh one or more processors. Accordingly, as an example, '~' indicates components such as software components, object-oriented software components, class components, and task components, and processes, functions, properties, and procedures. , subroutines, segments of program code, drivers, firmware, microcode, circuitry, data structures and variables. The functions provided in the components and '~ units' may be combined into a smaller number of components and '~ units' or further separated into additional components and '~ units'.

Hereinafter, with reference to the accompanying drawings, a preferred embodiment of a method for searching a real estate area and for sale based on a personal residential value evaluation standard using artificial intelligence according to the present invention will be described in detail.

First, a description will be given of a real estate area and property search apparatus based on personal residential value evaluation criteria using artificial intelligence according to a preferred embodiment of the present invention with reference to FIG. 1 .

1 is a block diagram for explaining a real estate area and property search apparatus based on personal residential value evaluation criteria using artificial intelligence according to a preferred embodiment of the present invention.

Referring to FIG. 1 , a real estate area and sale search apparatus (hereinafter referred to as a 'real estate sale search apparatus') 100 based on personal housing value evaluation criteria using artificial intelligence according to a preferred embodiment of the present invention 100 is a user through a communication network. It is connected to the terminal 200 .

The real estate for sale search apparatus 100 searches for a real estate area and for sale that meet the user's individual housing value evaluation criteria using the user's preferred area description and the preferred sale description based on artificial intelligence.

Here, the criteria by which the user evaluates the housing value are as follows.

- Outdoor standard: indicates the characteristics of an individual's preferred area, that is, three zones. Here, three zones refer to areas with strengths in specific aspects, and in addition to school districts that indicate areas adjacent to stations, schools, and hagwons, which are generally considered to be high in residential value, the value may differ depending on the individual as a basis for selecting a residential area. Bungse-kwon, which indicates an area famous for bungeoppang, where a kickboard is available, a bicycle-seokwon that indicates an area with easy access to bicycles, a forested area that indicates an area where walking in the forest is easy, and a mall that indicates an area adjacent to a shopping mall It also includes Seulse-kwon, which indicates an area where facilities for wearing slippers and slippers can be accessed.

- Indoor standard: Indicates the interior structure of the property for sale (living room with river/forest view, soundproofing performance, spacious balcony, island kitchen, large dressing room, rooftop facility, etc.).

- Basic criteria: Basic information related to the property (apartment/multi-family/multi-family/single/neighborhood facility/share house/co-living house, etc.) presence, number of parking spaces, whether or not elevators are installed, price, size, etc.).

In addition, the preferred area description refers to a description of an area desired by the user according to the residential value evaluation standard, and may be in the form of a natural language text. Preferred property description refers to a description of the property desired by the user according to the residential value evaluation standard, and may be in the form of a natural language sentence. Of course, the description of the preferred area and the description of the property for sale may consist of keywords.

The user terminal 200 may access the real estate for sale search apparatus 100 through a communication network according to a user's operation to search for a real estate area and for sale to search for a real estate area and for sale.

Here, the user terminal 200 includes not only a desktop computer, but also a notebook computer, a workstation, a palmtop computer, an Ultra Mobile Personal Computer (UMPC), a tablet PC, a Personal Digital Assistant (PDA), a web pad, It can be made of a terminal having a memory means, such as a smart phone, a mobile phone, etc., and equipped with a microprocessor to have computational capability.

The communication network may include a telephone network as well as a data communication network including a local area network (LAN), a metropolitan area network (MAN), a wide area network (WAN), the Internet, etc. It does not discriminate between wired and wireless, and it does not matter which communication method is used.

Then, with reference to FIGS. 2 and 3, a method of searching for a real estate area and a property based on a personal residence value evaluation standard using artificial intelligence according to a preferred embodiment of the present invention will be described.

Figure 2 is a flow chart for explaining a real estate area and for sale search method based on personal residential value evaluation criteria using artificial intelligence according to a preferred embodiment of the present invention, Figure 3 is an individual using artificial intelligence according to a preferred embodiment of the present invention It is a diagram for explaining an example of a real estate area and property search process based on residential value evaluation criteria.

Referring to FIG. 2 , the real estate listing search apparatus 100 builds a regional (neighborhood) information corpus DB (S105).

In other words, the real estate sale search device 100 is an automatic information collection technique such as web crawling, a connection program, or a manual information collection technique from public or contracted information sources such as the Internet. Relevant unstructured natural language information can be collected. Here, the region-related atypical natural language information refers to information in various fields on a specific region, and there are reviews on restaurants, local introductions, academies, forests, trails, bicycle paths, road blockages, street food, and the like.

In addition, the real estate listing search apparatus 100 may process the collected region-related atypical natural language information to construct a regional information corpus DB.

More specifically, the real estate sale search device 100 analyzes the collected region-related unstructured natural language information for each region-related unstructured natural language information through an artificial intelligence natural language processing (NLP) technique, and builds a regional information corpus DB can do.

Here, the regional information corpus DB includes a regional item indicating a region such as an address, an alias entry indicating an alias of the corresponding region, an information input year item indicating the year in which information about the region was input, and a region description indicating information on the region It can contain items.

For example, the regional alias item may be acquired from region-related unstructured natural language information using a pre-constructed and stored lexicon. Here, the geographical name dictionary may consist of names (dong names, aliases, etc.) corresponding to each region. That is, by automatically recognizing a region name or region alias from the atypical natural language information related to the region through named entity recognition (NER), the relevant information may be added to the region description item corresponding to the region item.

In addition, the information input year item indicates the year in which the information was entered. If the region-related unstructured natural language information is public data, the date entered into the public data DB, if the region-related unstructured natural language information is a website, the blogging date; If there is a writing date input by the author of the article, it may be the corresponding writing date.

In addition, the region description item includes an information content item indicating the content of information about the region, a sentiment analysis result of the information content, that is, a sentiment analysis item indicating whether positive or negative content, and It may include an information source item indicating the source of the acquisition. Here, the sentiment analysis item may be obtained through a data collection process, a preprocessing (subjectivity detection) process, and a polarity detection process. The data collection process refers to the process of collecting region-related unstructured natural language information described above. The preprocessing process refers to a process of acquiring sentiment words expressing positive or negative emotions from texts (ie, region-related unstructured natural language information) collected using a pre-built and stored sentiment dictionary. Here, the sentiment dictionary may consist of words corresponding to positive (like, I love you, etc.) and words corresponding to negative (dislike, extremely dislike, etc.), and may be continuously updated. The polarity detection process refers to a process of acquiring sentiment analysis items corresponding to atypical natural language information related to a region based on the sentiment words acquired through the preprocessing process. In other words, in preparation for the frequency of appearance of emotional words expressing positivity and sensibility words expressing negation, based on emotional words that show a higher frequency of appearance, it is possible to obtain sentiment analysis items corresponding to atypical natural language information related to the region. can Of course, sentiment analysis items corresponding to region-related unstructured natural language information may be obtained by using a pre-trained and constructed emotional classification model of a neural network. In addition, the information source item indicates a source from which information is obtained, and may be a public data portal (www.data.or.kr), a website related to a neighborhood/housing, a blog, or an SNS.

Thereafter, the user terminal 200 inputs a description of a preferred area and a description of a property for sale by a user's manipulation to search for a real estate area and a property (S110).

Then, the user terminal 200 provides the preferred area description input by the user and the preferred sale description to the real estate sale search apparatus 100 (S115).

Then, the real estate listing search apparatus 100 searches for a plurality of regions having high regional similarity with the preferred region description using the region information corpus DB (S120), and provides the preferred region candidate search results to the user terminal 200 ( S125).

That is, the real estate property search device 100 analyzes semantic similarity through artificial intelligence natural language processing techniques based on the regional information corpus DB, the preferred region description, and the region with high regional similarity with the preferred region description. In order, a preferred region candidate search result may be provided to the user terminal 200 along with the regional similarity.

More specifically, for each region of the regional information corpus DB, the real estate sale search apparatus 100 provides a semantic similarity between the information content item and the preferred region description, the sentiment evaluation result according to the sentiment analysis item, the information input year item and the search year. Regional similarity can be obtained for each region by acquiring regional similarity based on the information depreciation obtained based on the information and the correction constant according to the information source item.

For example, the real estate listing search apparatus 100 may calculate the regional similarity with respect to the region through [Equation 1] below.

[Equation 1]

Regional similarity = ∑ (semantic similarity * sentiment analysis result * correction constant / information depreciation)

Here, the semantic similarity indicates the semantic similarity calculated based on each sentence describing the region in an information source such as a blog or the web and a preferred region description using a technique for determining the semantic similarity of texts based on conventional artificial intelligence natural language processing. . For example, text may be vectorized (digitized) and semantic similarity may be obtained using a conventional similarity calculation technique based on the vectorized text. Conventional similarity calculation techniques include a Jaccard similarity calculation method, a Euclidean similarity calculation technique, a Manhattan similarity calculation technique, a cosine similarity calculation technique, and a neural network-based similarity calculation method. Meanwhile, the text may be divided into detailed components such as words, morphemes, and grapheme units, and semantic similarity for each of the detailed components may be obtained, and the semantic similarity of the corresponding text may be obtained based on the semantic similarity for each of the detailed components.

And, the emotion analysis result is the emotion analysis result for the information content. If the content is positive, the value is set to "+1", if the content is negative, the value is set to "-1", and if it is neutral or not related, the value is set to "0".

In addition, the correction constant is acquired based on the reliability of the source of information acquisition and influence indicators identified through other system operation, etc., and the value of the correction constant is continuously analyzed through statistical analysis or reinforcement learning as data is accumulated. It is updated manually or automatically.

And, the information depreciation is a value representing the time from the creation time of the information to the current time, and is calculated through "search year - information input year".

Then, the real estate listing search apparatus 100 may obtain the preferred area description and the regional similarity by adding the similarity of the preferred area description and each sentence obtained as above for each region.

Then, the user terminal 200 displays the preferred area candidate search results provided from the real estate listing search apparatus 100 (S130). A preferred region is selected by the user's manipulation (S135).

Then, the user terminal 200 provides the selected preferred area to the real estate sale search apparatus 100 (S140).

Then, the real estate sale search apparatus 100 searches for a sale with a high degree of similarity with the preferred sale description using the real estate sale DB in the preferred area (S145), and provides the preferred sale candidate search result to the user terminal 200 (S150).

Here, the real estate for sale DB is real estate for sale basic information such as area, number of floors, rooms, bathrooms, transaction types, prices, etc. May include photos of the real estate for sale.

That is, the real estate sale search device 100 analyzes the similarity of the real estate sale description and the real estate sale photo through an artificial intelligence technique based on the real estate sale DB in which information about the real estate sale is stored, based on the selected preferred area, the preferred sale A preferred listing candidate search result may be provided to the user terminal 200 along with the similarity of the listing in the order of the description and the listing having a high similarity to the listing.

In more detail, the real estate for sale search apparatus 100 may search for a real estate for sale corresponding to a preferred area selected from the real estate for sale DB.

And, the real estate for sale search device 100, for each of the searched real estate for sale, the basic information similarity obtained based on the real estate sale basic information and the preferred sale description, the description similarity obtained based on the real estate sale description and the preferred sale description, and Based on the photo similarity obtained based on the real estate for sale photo and the preferred sale description, it is possible to obtain the similarity for each real estate for sale searched.

For example, the real estate for sale search apparatus 100 may calculate the similarity for sale with respect to one real estate for sale through [Equation 2] below.

[Equation 2]

Property similarity = (Basic information similarity * Correction weight 1) + (Description similarity * Correction weight 2) + (Photo similarity * Correction weight 3)

Here, the basic information similarity indicates the similarity calculated based on the basic information of the real estate sale and the preferred sale description by using the technique of determining the similarity of the text based on the conventional artificial intelligence natural language processing. For example, by extracting required items (floor area, number of rooms, etc.) from the description of the preferred sale through entity name recognition (NER), and analyzing whether the basic information of the real estate sale matches or not, the similarity can be calculated as the ratio of matching items to all required items. there is.

And, the descriptive similarity indicates the similarity calculated based on the description of the real estate sale and the preferred sale description by using the technique of determining the similarity of the text based on the conventional artificial intelligence natural language processing. That is, the descriptive similarity may be obtained by the same method as the technique for calculating the regional similarity in [Equation 1] above. However, information depreciation according to input time is not taken into account under the premise that only the real estate for sale is currently being traded.

And, the photo similarity indicates the similarity calculated based on the photo of the real estate sale and the preferred article description by using a technique for determining the similarity between the image and the text based on the conventional artificial intelligence image processing technique. For example, a real estate sale photo, that is, an image is preprocessed (brightness, color tone adjustment, size adjustment, etc.), and an object detection and localization operation is performed through a conventional object recognition algorithm based on the preprocessed image to detect the object in the preprocessed image, and the text corresponding to the object detected through image captioning window, island kitchen) and calculating the similarity of the text and basic information by calculating the similarity with the image text and matching ratio to the requirements included in the description of the preferred property obtained through the entity name recognition (NER) technique when analyzing the similarity of the acquired text and basic information can be obtained. Conventional object recognition algorithms include R-CNN series (Fast R-CNN, Faster-R-CNN, Mask R-CNN, etc.), R-FCN, Histogram of Oriented Gradients (HOG), Spatial Pyramid Pooling (SPP-net), These include Single Shot Detector (SSD), You Only Look Once (YOLO), and Blitznet. At this time, the matching value between the requirements of the description of the preferred property and the image text is not a one-to-one match of words (eg, living room window = living room window), but semantic similarity (eg living room window = large glass window in living room). It means exceeding a predetermined reference value (eg 70%). The reference value is corrected as data is accumulated.

Further, the correction weights 1 to 3 represent preset weights for each degree of similarity according to the reliability and importance of the information source. The correction weight is continuously updated manually or automatically through statistical analysis or reinforcement learning as data is accumulated.

Then, the user terminal 200 displays the search result of the preferred sale candidate provided from the real estate sale search apparatus 100 (S155).

In addition, the user terminal 200 may select one preferred item from among the preferred item candidate search results (S160).

On the other hand, the real estate area and property search method based on the personal residential value evaluation criteria according to the present invention search for and provide a preferred area candidate based on the preferred area description input by the user, and in the preferred area selected from the preferred area candidates, 2 and 3 are shown in FIGS. 2 and 3 as "searching for and providing a preferred item for sale to the user based on the preferred item for sale description input by the user", the present invention is not limited thereto, and according to an embodiment, "a preferred item inputted by the user is provided" It is also possible to search for and provide a preferred listing candidate based on the description, and search for a preferred area candidate based on the preferred area description input by the user from the preferred property selected from among the preferred listing candidates and provide it to the user. Of course, according to the present invention, "searching for a preferred area candidate based on the preferred area description and providing it to the user" and "searching for a preferred item for sale candidate based on the preferred item description and providing it to the user" may be simultaneously performed.

Then, an example of a real estate area and property search process based on personal housing value evaluation criteria using artificial intelligence according to a preferred embodiment of the present invention will be described with reference to FIGS. 4 to 6 .

4 is a diagram illustrating an example of a real estate area and property search process based on personal residential value evaluation criteria using artificial intelligence according to a preferred embodiment of the present invention.

Referring to FIG. 4 , a regional information corpus DB is built in advance using region-related atypical natural language information collected through a web crawling technique from public information sources such as the Internet. In addition, a real estate for sale DB consisting of information on real estate for sale is built in advance.

As shown in FIG. 4 , when a preferred area description and a preferred sale description are input by the user, the real estate sale search apparatus 100 uses the area information corpus DB to select a region (neighborhood) with a high regional similarity to the preferred area description It is possible to search and provide recommended neighborhoods (preferred local candidates) to the user. In addition, the real estate listing search apparatus 100 may search for a listing having a high similarity to the preferred listing description and listing using the real estate listing DB, and may provide a recommended listing (preferred listing candidate) to the user.

In this case, the real estate sale search apparatus 100 provides the user with a preferred area candidate searched based on the preferred area description, and in a preferred area selected by the user from among the preferred area candidates, the user selects the preferred item for sale candidate searched based on the preferred area description. can be provided to On the contrary, the real estate sale search apparatus 100 provides the user with the preferred sale candidate searched based on the preferred sale description, and in the preferred sale selected by the user from among the preferred sale candidates, the preferred area candidate searched based on the preferred area description can be provided to users. Of course, the real estate for sale search apparatus 100 may simultaneously provide the preferred area candidate based on the preferred area description to the user and the preferred item for sale candidate to the user based on the preferred item for sale description.

FIG. 5 is a diagram illustrating in more detail the preferred area search process shown in FIG. 4 .

The real estate sale search apparatus 100 analyzes regionally-related unstructured natural language information collected through a web crawling technique from public information sources such as the Internet for each region-related atypical natural language information through an artificial intelligence natural language processing technique, as shown in FIG. It is possible to build a local information corpus DB such as a bar. For example, by preprocessing sentences collected from websites, blogs, etc., special sentences and stopwords may be deleted to collect regional (neighborhood) descriptions (corpus). Then, the region name can be automatically recognized through entity name recognition (NER), and related information can be added to the region description item of the corresponding region. In this case, the emotion analysis result of the information may also be reflected. For example, "a disadvantage of difficult parking" indicates negative contents related to parking, and "like" indicates positive contents.

And, the real estate sale search apparatus 100 analyzes the similarity of sentences through natural language processing (NLP) based on the preferred area description (preferred area description input by the user) and the area information corpus DB to obtain regional similarity by region, Based on the obtained regional similarity, a search result of a preferred region candidate may be provided to the user.

6 is a view showing in more detail the process of searching for a preferred item shown in FIG. 4 .

Real estate listing basic information, such as area, number of floors, rooms, bathroom, transaction type, price, etc. Using the real estate for sale DB as shown in FIG. 6, which consists of real estate for sale information, the real estate for sale search apparatus 100 searches for a preferred sale candidate based on the preferred sale description (preferred sale description input by the user) to find the preferred sale Candidates can be provided to users.

That is, the real estate listing search apparatus 100 may calculate the basic information similarity by analyzing whether the real estate listing basic information and the preferred listing description match through entity name recognition (NER).

In addition, the real estate sale search apparatus 100 may calculate the description similarity by analyzing the similarity between the preferred sale description and the real estate sale description through natural language processing (NLP).

And, the real estate sale search apparatus 100 identifies whether the elements (eg, island kitchen, living room window, etc.) are included in the preferred sale description through object detection and image captioning from the real estate sale photo, that is, the image. Thus, the photo similarity can be calculated.

Then, the real estate for sale search device 100 calculates the similarity for sale with the preferred sale description of the corresponding real estate based on the basic information similarity, description similarity, and photo similarity, and selects a preferred sale candidate based on the similarity for sale by real estate sale can be provided to

Even though all the components constituting the embodiment of the present invention described above are described as being combined or operated in combination, the present invention is not necessarily limited to this embodiment. That is, within the scope of the object of the present invention, all the components may operate by selectively combining one or more. In addition, all of the components may be implemented as one independent hardware, but a part or all of each component is selectively combined to perform some or all of the functions of the combined hardware in one or a plurality of hardware program modules It may be implemented as a computer program having In addition, such a computer program is stored in a computer readable media such as a USB memory, a CD disk, a flash memory, etc., read and executed by a computer, thereby implementing an embodiment of the present invention. The recording medium of the computer program may include a magnetic recording medium, an optical recording medium, and the like.

The above description is merely illustrative of the technical idea of the present invention, and those of ordinary skill in the art to which the present invention pertains may make various modifications, changes and substitutions within the scope without departing from the essential characteristics of the present invention. will be. Accordingly, the embodiments disclosed in the present invention and the accompanying drawings are for explaining, not limiting, the technical spirit of the present invention, and the scope of the technical spirit of the present invention is not limited by these embodiments and the accompanying drawings . The protection scope of the present invention should be construed by the following claims, and all technical ideas within the scope equivalent thereto should be construed as being included in the scope of the present invention.

100: real estate sale search device,
200: user terminal

Claims (4)

Although it may not be considered in the normal housing value evaluation criteria using artificial intelligence, in order to search real estate areas and properties based on the housing value evaluation that is important to individuals
collecting region-related unstructured natural language information through a web crawling technique from published information sources, and processing the collected unstructured natural language information to construct a regional information corpus DB;
receiving a preferred area description and a preferred sale description for the real estate area and sale through the user terminal;
Based on the regional information corpus DB, based on the preferred region description, semantic similarity is analyzed through artificial intelligence natural language processing technique, and the preference region description and the region with the highest regional similarity are preferred together with regional similarity in the order of providing a local candidate search result to the user terminal;
receiving a preferred area selected from the preferred area candidate search result through the user terminal; and
Based on the real estate for sale DB in which information about real estate for sale is stored, in the selected preferred area, the similarity of the real estate sale description and real estate sale photo is analyzed through artificial intelligence techniques, and the preferred sale description and the sale with high similarity are in the order of providing, to the user terminal, a preferred listing candidate search result together with the similarity of the listing;
A method of searching real estate areas and properties based on individual housing value evaluation criteria using artificial intelligence, including In claim 1,
The step of constructing the local information corpus DB is,
By analyzing the collected unstructured natural language information for each unstructured natural language information through an artificial intelligence natural language processing technique, the regional information corpus DB including a regional item, an alias item, an information input year item, and a regional description item consists of constructing,
A method of searching real estate areas and properties based on individual housing value evaluation criteria using artificial intelligence. In claim 2,
The step of providing a search result for a candidate for a preferred area includes:
For each region of the regional information corpus DB, the semantic similarity between the information content item of the region description item and the preferred region description, the emotion evaluation result according to the sentiment analysis item of the region description item, the information input year item and the search year Comprising of acquiring regional similarity by region by acquiring regional similarity based on the correction constant according to information depreciation obtained based on the information source reliability of the regional description item, etc.,
A method of searching real estate areas and properties based on individual housing value evaluation criteria using artificial intelligence. In claim 1,
The step of providing the search result of the candidate for the preferred property,
Search for real estate for sale corresponding to the selected preferred area in the real estate for sale DB, and for each of the searched real estate for sale, basic information for sale on the basis of the basic information for sale and the preferred sale description similarity, real estate sale description and the preference Based on the similarity of the description obtained based on the description for sale and the photos and drawings for sale of real estate and the similarity of the photos obtained based on the description for the preferred sale, it consists of obtaining the similarity for sale for each searched real estate for sale,
A method of searching real estate areas and properties based on individual housing value evaluation criteria using artificial intelligence.

Images