Semantic information retrieval for augmented reality domain

Abstract

A method and apparatus for semantic information retrieval for an augmented reality domain are disclosed. The method for semantic information retrieval for an augmented reality domain may include the step of performing semantic information retrieval in an AR domain using an AR ontology composed of an augmented reality concept.

Description

Technology Field

[0001] The following description concerns technology for accessing information in an augmented reality environment. Background Technology

[0002] Augmented reality (AR) combines digital information (text, images, videos, 3D models, etc.) with reality, and as a future user interface, AR plays an important role in supporting both consumer and industrial applications.

[0003] Research and development (R&D) of AR technology is increasing in various application fields such as learning, health, tourism, and manufacturing. For example, Korean registered patent No. 10-2178894 (registered on November 9, 2020) discloses a technology that provides augmented reality-based learning content using an image printed on a mat.

[0004] With the recent emergence of the metaverse, the popularity of AR is rising, and accessing concise, accurate, and precise information in this field is becoming difficult on the World Wide Web (WWW).

[0005] In relation to accessing correct information through search engines, semantic information retrieval through semantic analysis delivers more relevant information related to the user's query.

[0006] Although various methods for semantic information retrieval are being developed, research on semantic information retrieval methods in the AR domain is lacking. The problem to be solved

[0007] To perform effective information retrieval from web documents, we provide an AR search engine based on semantic information retrieval methods in the AR domain.

[0008] We provide an AR search engine that automatically organizes, understands, searches, and summarizes web documents to improve relevance scores in the AR domain.

[0009] It provides an AR ontology for clustering AR documents into AR topics and concepts.

[0010] An ontology-based clustering method is provided to explore and cluster AR documents using an ontology, a k-means clustering algorithm, a vector space model, and a TF-IDF (term frequency-inverse document frequency) weighting model. means of solving the problem

[0011] A method for retrieving semantic information of a computer device comprising at least one processor, wherein the method comprises the step of performing semantic information retrieval in an AR domain using an AR ontology composed of an AR concept by the at least one processor.

[0012] According to one aspect, the AR ontology may include hardware, software, tracking, interaction, and interface related to AR.

[0013] According to another aspect, the step of performing the above may include classifying web crawler results corresponding to a search query into AR fields through the AR ontology.

[0014] According to another aspect, the step of performing the above may include the step of clustering AR documents into AR-related topics or concepts using the AR ontology.

[0015] According to another aspect, the steps performed above may include representing AR documents as vectors according to the concept of the AR ontology; and clustering the AR document vectors into the AR domain.

[0016] According to another aspect, the above-mentioned representing step can represent the AR document as an N-dimensional vector using a semantic vector space model (SVSM).

[0017] According to another aspect, the above-described representing step may include representing the AR document as an N-dimensional vector using a semantic vector space model (SVSM), creating a 2-dimensional (M×N) matrix containing M concepts for each document; assigning weights to each concept based on a term frequency-inverse document frequency (TF-IDF) score representing the relationship between concepts corresponding to each pair of the 2-dimensional matrix; and generating an SVSM for the document by applying the maximum weight to the superclass of the seed concept.

[0018] According to another aspect, the clustering step may include the step of clustering the AR document vectors using k-means clustering to display the information retrieval results as clusters.

[0019] According to another aspect, the step of performing the above may further include a step of preprocessing to convert the query and AR document into a word sequence prior to the step of expressing the above.

[0020] According to another aspect, the preprocessing step described above may be performed through at least one of the following: a process of performing tokenization, which involves breaking down the query and AR document into sentences and then into individual words or terms, or a process of removing stop-words included in the query and AR document.

[0021] A computer program stored on a computer-readable recording medium is provided to execute the above-mentioned semantic information retrieval method on the computer device.

[0022] A computer device is provided that includes at least one processor implemented to execute a readable command on a computer device, wherein the at least one processor performs semantic information retrieval in an AR domain using an AR ontology composed of an augmented reality concept. Effects of the invention

[0023] According to embodiments of the present invention, by providing an AR search engine based on a semantic information retrieval method in an AR domain to perform effective information retrieval in web documents, users can obtain more relevant results for search queries of different groups and obtain very fast results in the AR domain. Brief explanation of the drawing

[0024] FIG. 1 is a block diagram illustrating an example of the internal configuration of a computer device in an embodiment of the present invention. FIG. 2 illustrates the major classes of an AR ontology in one embodiment of the present invention. FIG. 3 illustrates an AR software subclass in one embodiment of the present invention. FIG. 4 illustrates an AR hardware subclass in one embodiment of the present invention. FIG. 5 illustrates AR tracking, interface, and interaction subclasses in one embodiment of the present invention. FIG. 6 illustrates a methodology of an ontology-based AR search engine in one embodiment of the present invention. FIG. 7 is a diagram illustrating the Semantic Vector Space Model (SVSM) in an embodiment of the present invention. FIG. 8 shows pseudocode of a pseudo information retrieval method for an AR search engine in one embodiment of the present invention. Specific details for implementing the invention

[0025] Hereinafter, embodiments of the present invention will be described in detail with reference to the attached drawings.

[0027] Embodiments of the present invention relate to a technology for accessing information in an augmented reality environment.

[0028] Embodiments including those specifically disclosed in this specification can improve relevance scores in the AR domain through an AR search engine that automatically organizes, understands, searches, and summarizes web documents, and allow users to organize and manage relevant AR documents by various AR concepts and discover more accurate results quickly and efficiently in terms of relevance in the AR field.

[0029] A semantic information retrieval device according to embodiments of the present invention may be implemented by at least one computer device, and a semantic information retrieval method according to embodiments of the present invention may be performed through at least one computer device included in the semantic information retrieval device. At this time, a computer program according to an embodiment of the present invention may be installed and run on the computer device, and the computer device may perform a semantic information retrieval method according to embodiments of the present invention under the control of the run computer program. The above-described computer program may be stored on a computer-readable recording medium to be combined with the computer device to execute the semantic information retrieval method on the computer.

[0030] FIG. 1 is a block diagram illustrating an example of a computer device according to an embodiment of the present invention. For example, a semantic information retrieval device according to embodiments of the present invention may be implemented by a computer device (100) illustrated in FIG. 1.

[0031] As illustrated in FIG. 1, a computer device (100) may include a memory (110), a processor (120), a communication interface (130), and an input / output interface (140) as components for executing a semantic information retrieval method according to embodiments of the present invention.

[0032] Memory (110) is a computer-readable recording medium and may include a non-perishable mass storage device such as RAM (random access memory), ROM (read only memory), and a disk drive. Here, a non-perishable mass storage device such as a ROM and a disk drive may be included in the computer device (100) as a separate permanent storage device distinct from memory (110). Additionally, an operating system and at least one program code may be stored in memory (110). These software components may be loaded into memory (110) from a computer-readable recording medium separate from memory (110). This separate computer-readable recording medium may include a computer-readable recording medium such as a floppy drive, disk, tape, DVD / CD-ROM drive, or memory card. In another embodiment, software components may be loaded into memory (110) through a communication interface (130) rather than a computer-readable recording medium. For example, software components can be loaded into the memory (110) of the computer device (100) based on a computer program installed by files received through the network (160).

[0033] The processor (120) may be configured to process instructions of a computer program by performing basic arithmetic, logic, and input / output operations. Instructions may be provided to the processor (120) via memory (110) or a communication interface (130). For example, the processor (120) may be configured to execute instructions received according to program code stored in a recording device such as memory (110).

[0034] The communication interface (130) may provide a function for the computer device (100) to communicate with other devices through a network (160). For example, requests, commands, data, files, etc. generated by the processor (120) of the computer device (100) according to program code stored in a recording device such as memory (110) may be transmitted to other devices through the network (160) under the control of the communication interface (130). Conversely, signals, commands, data, files, etc. from other devices may be received by the computer device (100) through the communication interface (130) of the computer device (100) via the network (160). Signals, commands, data, etc. received through the communication interface (130) may be transmitted to the processor (120) or memory (110), and files, etc. may be stored in a storage medium (the permanent storage device described above) that the computer device (100) may further include.

[0035] The communication method is not limited and may include not only communication methods utilizing communication networks that the network (160) may include (e.g., mobile communication networks, wired internet, wireless internet, broadcasting networks, etc.), but also short-range wired / wireless communication between devices. For example, the network (160) may include any one or more networks such as a PAN (personal area network), LAN (local area network), CAN (campus area network), MAN (metropolitan area network), WAN (wide area network), BBN (broadband network), and the Internet. Additionally, the network (160) may include any one or more network topologies such as a bus network, star network, ring network, mesh network, star-bus network, tree or hierarchical network, but is not limited thereto.

[0036] The input / output interface (140) may be a means for interfacing with an input / output device (150). For example, the input device may include a device such as a microphone, keyboard, camera, or mouse, and the output device may include a device such as a display or speaker. As another example, the input / output interface (140) may be a means for interfacing with a device in which the functions for input and output are integrated into one, such as a touchscreen. The input / output device (150) may be composed of a computer device (100) and a single device.

[0037] Additionally, in other embodiments, the computer device (100) may include fewer or more components than the components of FIG. 1. However, it is not necessary to clearly illustrate most of the prior art components. For example, the computer device (100) may be implemented to include at least some of the input / output devices (150) described above, or may include other components such as a transceiver, a camera, various sensors, a database, etc.

[0038] Below, specific embodiments of semantic information retrieval technology in the augmented reality domain will be described.

[0039] In this embodiment, an AR search engine based on a semantic information retrieval method in the AR domain is provided to perform effective information retrieval from web documents. Through this, users can obtain more relevant results for search queries of different groups and obtain very fast results in the AR domain.

[0040] To this end, the present invention provides an AR ontology that calculates relevance scores in semantic information retrieval by examining key AR concepts such as software, hardware, tracking, interaction, and interface. Furthermore, the present invention provides semantic information retrieval based on an AR ontology and a clustering method that performs semantic analysis of web documents, retrieves syntactically and semantically relevant information therefrom, and clusters it according to AR concepts. This is a query result clustering problem performed to improve the accuracy of information retrieval.

[0041] The main content of the present invention is an AR ontology for clustering AR web documents into AR topics and an ontology-based clustering method for an information retrieval system, and the main advantages are fast search time and semantic search through the hierarchical inheritance of the ontology.

[0042] First, we will explain the AR ontology to be used for semantic information retrieval in the AR domain, as well as the superclasses, subclasses, and their relationships based on the AR concept.

[0043] AR Ontology

[0044] AR ontologies are constructed in four stages: specification, knowledge acquisition / conceptualization, implementation, and evaluation.

[0045] (1) AR Ontology Specification

[0046] The ontology specification is the first step in the ontology development process, determining the scope, purpose, intended users, and requirements of the AR ontology. The purpose of the AR ontology is to support semantic information retrieval from web documents by providing a structured representation of augmented reality concepts; it is a shared and reusable representation of AR conceptualizations. The scope of the AR ontology includes representations of AR and development-related entities for various applications, based on AR definitions and characteristics. The end users of the AR ontology are developers, researchers, stakeholders, and relevant authorities. Developers can directly use the AR ontology to develop various applications, including search engines and AR information retrieval. Other end users can indirectly use the ontology through software and applications developed based on it.

[0047] AR includes a ubiquitous user interface with three characteristics that aims to visualize digital / virtual information directly registered in the real world. An AR system must combine the real and virtual worlds, including real-time interaction, and is registered in 3D. The definition of AR requires specific hardware / devices, spatial registration, and interaction. Virtual information must dynamically align the tracking system by dynamically determining spatial attributes. Intuitive interaction between the user and virtual information is created through the appropriate user. Finally, it is software that enables the hardware to perform the tasks required by the AR system, regardless of the hardware platform.

[0048] (2) Knowledge acquisition and conceptualization

[0049] This step involves reviewing relevant knowledge regarding AR technology concepts found through search engines for the development of AR systems in various applications. Relevant knowledge regarding AR ontologies is obtained from existing ontologies and existing research. Existing ontologies do not encompass all AR concepts and are mostly developed for specific applications. Therefore, research focused on AR concepts is used to develop AR ontologies. After collecting knowledge from existing ontologies and related research, relevant terms are listed to define class hierarchies, class properties, and individuals. Based on the definition of AR, five superclasses are defined: hardware, software, interaction, tracking, and interface. Relevant classes are defined as subclasses of superclasses using the 'SubClass Of' attribute. Figure 2 illustrates the major classes of the AR ontology based on the aforementioned concepts.

[0050] (3) AR Ontology Implementation

[0051] AR ontologies can be implemented in the OWL format in Protege software, and OWL files can be utilized in the semantic information retrieval process for AR search engine development. OWL, which includes classes, object attributes, and data attributes, is used to cluster web AR documents.

[0052] (4) AR Ontology Evaluation

[0053] There are several approaches for evaluating AR ontologies. Task-based or application-based evaluation approaches are applied to evaluate AR ontologies. This approach assesses the performance of an ontology within the context of an application by measuring its capability to complete target tasks. AR ontologies are evaluated in semantic information retrieval systems.

[0054] Ontology Classes and Individuals

[0055] Software is related to creating and using augmented reality systems that can be directly used in AR applications, creating AR applications, or providing content to AR applications.

[0056] As illustrated in Fig. 3, software can be divided into two main classes: authoring tools and development tools. Authoring tools enable users with no programming background to create virtual content using AR. There are various types of AR authoring tools that can be used to create augmented reality experiences, one of which is DART (Designers AR Toolkit). Development tools include Software Development Kits (SDKs), game engines, design / modeling software, languages, and libraries. An AR SDK is a collection of development tools that provides various functions, such as AR recognition, AR tracking, and AR content rendering. SDKs reduce integration and development time for software professionals. Vuforia, Wikitude, ARCore, and ARToolkit are examples of widely used AR SDKs for developing AR applications. The recognition and tracking capabilities of the Vuforia engine can support different types of objects in both 2D and 3D. The Vuforia SDK supports both Android and iOS platforms. Horizontal surfaces, such as floors and the top surface of tablets, can also be detected and tracked. Game engines can interact with AR sensors and displays by providing an integrated AR framework. Information to be sent to a renderer to generate signals for display devices is generated by a game engine, such as the Unity development tool. A key feature of a game engine is that it provides a simulation loop that supports user interaction.

[0057] Figure 4 illustrates AR hardware classes and their individuals. The primary AR devices are displays and input devices. Displays transmit signals to the visual, tactile, auditory, and olfactory senses, as well as, in some cases, the gustatory sense. Therefore, both visual and non-visual displays are included in the display class. Non-visual displays include auditory, tactile, olfactory, and gustatory displays. The primary function of a visual display is to generate light signals that the eye interprets as visual images. In AR, there are four types of visual displays: HMDs (head-mounted displays) / HADs (head-attached displays), HUDs (head-up displays), handheld displays, and spatial displays. In addition to displays, input devices help create immersive environments where users can interact with virtual content, including selecting, manipulating, and creating virtual content. Vive controllers, gloves, pointing devices, joysticks, and touchpads are examples of input devices.

[0058] HMDs, which overlay both real and virtual images within the user's worldview, utilize the device either on the head or as part of a helmet. The merging of real and virtual images can be achieved by using video-view or optical-view HMD systems. Video-view HMDs require two cameras and their processing to provide mismatched resolutions for both the "real parts" of the augmented scene and virtual objects, whereas optical-view HMDs use half-silver mirror technology, allowing a view of the real world to pass through lenses while graphically overlaid information is reflected into the user's eyes. A Head-Up Display (HUD) is a method of projecting information onto the windscreen, which is widely used in vehicles. HUDs provide a transparent screen that visualizes information without the driver having to take their eyes off the road. This type of display integrates information about the external environment into the windshield screen, allowing the driver to focus their attention in a new way. Portable devices such as smartphones, tablets, and PDAs are small computing devices equipped with a type of panel for displaying on a mobile projector. They support mobile AR applications because they possess sufficient computing power, and the projector is small and portable enough. Finally, spatial displays augment digital information onto physical objects, enabling AR experiences without the need to wear or carry a display. This type of display supports user grouping and collaboration among users.

[0059] The tracking class, one of the most important classes of AR systems, includes a system that performs virtual object registration in the real environment. The tracking system aims to accurately align virtual information with the real world in real time to provide an engaging AR experience. Consequently, the tracking system is utilized to dynamically identify the spatial properties (6DOF) of virtual objects (three components for position and three components for orientation) in the real world.

[0060] There are various methods for AR tracking, which are classified into marker-based methods and markerless methods. Referring to Fig. 5, marker-based methods rely on physical markers, such as tags and reference markers, to identify objects in the real world. Markers include unique visual characteristics, such as many corners and edges, which can be easily detected by computer vision techniques. Markerless methods have two subclasses. The former includes tracking methods that use computer vision and image processing techniques to track AR users based on natural characteristics that can be easily detected in the scene. The latter is a tracking method that estimates the position and orientation of the camera by analyzing sensor data.

[0061] Finally, AR interfaces and interactions constitute two major classes in AR systems. AR interfaces apply various technologies to create intuitive interactions between the user and digital content. AR interfaces rely on the number and variety of information inputs and outputs that enable the user to interact with the AR system. Tangible AR interfaces, collaborative AR interfaces, shared AR interfaces, haptic AR interfaces, natural interfaces, and multi-modal interfaces are examples of AR interfaces. Interaction methods include gesture recognition, facial recognition, speech recognition, touching, pointing, and grabbing.

[0062] Next, we will describe a methodology based on a vector space model based on the k-means clustering algorithm concept and a TF-IDF approach for AR search engines.

[0063] In the AR field, existing web crawler results are failing to meet efficiency and effectiveness requirements, even when using advanced document indexing approaches, due to the exponential increase in the volume of available information. For example, while searching for which AR SDKs are being utilized in the agricultural sector, users typically search using different keywords, such as "AR in agriculture." If search keywords are used alone, the crawler provides irrelevant results; instead, the crawler must classify the results to highlight those related to the AR SDKs of interest. Furthermore, multiple keywords must be used to search for desired results, and classifying web crawler results by the AR field allows for the efficient acquisition of desired findings.

[0064] The primary objective of the AR search engine according to the present invention is to effectively analyze user queries in an AR concept with minimal effort and time. To integrate semantics into comprehensive web search tools, existing tools can be enriched using the aforementioned AR ontology.

[0065] Figure 6 illustrates a methodology for an ontology-based AR search engine. The methodology according to the present invention consists of three stages: pre-processing, SVSM (Semantic Vector Space Modeling), and clustering. In the pre-processing stage, user queries and web documents are analyzed to generate data in a clean and uniform format for the following processes. The SVSM stage involves representing AR documents as vectors according to the concept of an AR ontology. The clustering stage is performed by semantically clustering all web documents into AR domains.

[0066] (1) Preprocessing

[0067] A preprocessing step is necessary to obtain efficient results. This step is performed to convert queries and AR documents into sequences of words using Natural Language Processing (NLP) techniques, including tokenization and stop-word removal. Tokenization is the process of breaking down input queries and AR documents into sentences and then into individual words or terms. Stop-word lists are used to remove all non-informative terms.

[0068] (2) SVSM stage

[0069] The SVSM (semantic vector space model) is used in the modeling phase. AR Document A i is A i ={w n It is represented as an N-dimensional vector denoted as { , n=1, ..., N}. Here, N is the number of main concepts of the AR ontology and w n is the weight of each concept. To create an SVSM, it is assumed that ontology elements, including classes or individuals, represent concepts in the AR domain. The key concepts are the superclass and clustering used to represent each AR document. The vector for each document is generated in three steps. First, D={A i Each document denoted as { , i-1, ..., L} forms a 2-dimensional (M×N) matrix containing M concepts. Here, each element of the matrix is ​​a weight w corresponding to the superclass of the seed concept. mn It represents. Second, for each pair of m and n, s mn The concept indicated by and t m Using the semantic similarity between the TF-IDF scores of the m-th concept denoted by w mn Calculate. Fig. 7 illustrates the matrix and SVSM along with an example of an AR ontology domain. In Fig. 7 Use to represent key concepts.

[0070] [Mathematical Formula 1]

[0071]

[0072] Semantic similarity s mn is the shortest distance d between concepts in an AR ontology. mn It is determined inversely proportional to, and according to the ontology hierarchy, the distance between two concepts is the number of links between the two concepts. For example, the distance between GPS and tracking is 3, and in the AR ontology, a lower similarity score means that the distance between the two concepts is greater. Therefore, semantic similarity is mathematically defined as follows.

[0073] [Mathematical Formula 2]

[0074]

[0075] The TF-IDF score t m is each AR document A i It indicates the importance of concepts within. In this embodiment, weights are assigned to concepts using the TF-IDF technique, which investigates the relationships between each concept within a document collection. This technique extracts keywords from a document collection in several stages, such as calculating similarity between documents and determining search rankings. TF measures the frequency of occurrence of a concept to reflect its importance within a document, and a high TF value indicates that the concept has high importance within the document. The importance of a concept within a document collection is measured by IDF, which is the inverse of the number of times the concept appears.

[0076] [Mathematical Formula 3]

[0077]

[0078] [Mathematical Formula 4]

[0079]

[0080] [Mathematical Formula 5]

[0081]

[0082] Here, is concept C mThis AR document A i It is the number of times it appeared in, is concept C m This represents the number of AR documents that appeared. After creating the matrix in the second step, the SVSM for each AR document can be completed in the third step by applying the maximum weight for the given superclass.

[0083] [Mathematical Formula 6]

[0084]

[0085] (3) Clustering step

[0086] Clustering, which creates clusters by grouping similar terms extracted from documents, is a type of text mining process. In this step, k-means clustering is used to cluster AR document vectors and display information retrieval results as clusters. The k-means algorithm is one of the most widely used unsupervised learning algorithms and has the following objective function that must be minimized during an iterative process through necessary conditions.

[0087] [Mathematical Formula 7]

[0088]

[0089] Here, is the objective function, B={b k} is a K cluster center, and is A as a binary value i It is 1 if it belongs to the k-th cluster. In the classical K-means algorithm is AR document A i and cluster center b k It is the Euclidean distance between them. When applying the k-means algorithm to clustered AR documents, cosine distance is used, which provides better results when finding text similarity. Two documents of D (A p ∈D and A qAssuming ∈D, where p and q are the document indices, the cosine distance between two AR documents is calculated using the following formula. class are A respectively p Document and A q Key concepts regarding the document This is the weight calculated for.

[0090] [Mathematical Formula 8]

[0091]

[0092] [Mathematical Formula 9]

[0093]

[0094] These embodiments enable fast search times and semantic search through the hierarchical inheritance of ontologies. In other words, the AR search engine can improve search performance in the AR domain with the least amount of effort and time because it provides a simpler and more efficient tool for all users to obtain essential information by minimizing search keywords.

[0095] To evaluate AR search engines, which are AR search engines in the AR domain, two evaluation metrics can be used: task completion time and mean reciprocal rank (MRR). Task completion time measures the efficiency of a search engine and refers to the total time required for a participant to find the desired result in a search. MRR measures the average of the inverse of the multiplicative ranks of the correct result in the sorted list of output results from the search engine. MRR refers to how many results a participant checks until they obtain the desired result and can be calculated using Equation 10.

[0096] [Mathematical Formula 10]

[0097]

[0098] Here, IQI is the number of searches, q is the search index, and rank qrepresents the rank position of the first correct result for the qth search.

[0099] Figure 8 shows the pseudocode of a pseudo-information retrieval method for an AR search engine.

[0100] As such, according to the embodiments of the present invention, semantic information retrieval based on an AR ontology that calculates a relevance score in semantic information retrieval by examining key concepts of AR such as software, hardware, tracking, interaction, and interface can be provided, and a method for performing semantic analysis on web documents for semantic information retrieval and searching for syntactic and semantic-related information to cluster according to AR concepts can be provided. Through the AR search engine according to the present invention, users can organize and manage related AR documents by various AR concepts, and can discover more accurate results quickly and efficiently in terms of relevance in the AR field.

[0101] The device 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 device and components described in the embodiments may be implemented using one or more general-purpose or special-purpose computers, such as a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable gate array (FPGA), a programmable logic unit (PLU), a microprocessor, or any other device capable of executing and responding to instructions. The processing unit may execute an operating system (OS) and one or more software applications executed on said operating system. Additionally, the processing unit may access, store, manipulate, process, and generate data in response to the execution of the software. For ease of understanding, the processing unit may be described as being used as a single unit, but those skilled in the art will understand that the processing unit may include multiple processing elements and / or multiple types of processing elements. For example, the processing unit may include multiple processors or one processor and one controller. In addition, other processing configurations, such as parallel processors, are also possible.

[0102] Software may include computer programs, code, instructions, or a combination of one or more of these, and may configure a processing unit to operate as desired or instruct the processing unit independently or collectively. Software and / or data may be embodied in any type of machine, component, physical device, computer storage medium, or device so as to be interpreted by the processing unit or to provide instructions or data to the processing unit. Software may be distributed over networked computer systems and may be stored or executed in a distributed manner. Software and data may be stored on one or more computer-readable recording media.

[0103] 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 on a computer-readable medium. In this case, the medium may continuously store a program executable by a computer, or temporarily store it for execution or download. Additionally, the medium may be various recording or storage means in the form of a single or several hardware combined, and may not be limited to a medium directly connected to a computer system but may exist distributed over a network. Examples of media may include magnetic media such as hard disks, floppy disks, and magnetic tapes; optical recording media such as CD-ROMs and DVDs; magneto-optical media such as floptical disks; and media configured to store program instructions, including ROM, RAM, and flash memory. Additionally, other examples of media may include recording or storage media managed by app stores that distribute applications or sites and servers that supply or distribute various other software.

[0104] Although the embodiments have been described above with reference to limited examples and drawings, those skilled in the art can make various modifications and variations from the description above. For example, suitable results can be achieved even if the described techniques are performed in a different order than described, and / or the components of the described system, structure, device, circuit, etc. are combined or assembled in a form different from described, or replaced or substituted by other components or equivalents.

[0105] Therefore, other implementations, other embodiments, and equivalents to the claims also fall within the scope of the claims set forth below.

Claims

Claim 1 A method for retrieving semantic information of a computer device comprising at least one processor, wherein the at least one processor performs semantic information retrieval in an AR domain using an AR ontology composed of an AR concept (augmented reality concept), and the step of performing the semantic information retrieval includes the step of representing an AR document as a vector using a semantic vector space model (SVSM) according to the concept of the AR ontology. Claim 2 A semantic information retrieval method according to claim 1, characterized in that the AR ontology includes hardware, software, tracking, interaction, and interface related to AR. Claim 3 A semantic information retrieval method according to claim 1, wherein the step of performing the above includes the step of classifying web crawler results corresponding to a search query into AR fields through the AR ontology. Claim 4 A semantic information retrieval method according to claim 1, wherein the step of performing the above includes the step of clustering AR documents into AR-related topics or concepts using the AR ontology. Claim 5 A semantic information retrieval method according to claim 1, wherein the step of performing the above further includes the step of clustering AR document vectors into the AR domain. Claim 6 delete Claim 7 A semantic information retrieval method according to claim 1, wherein the representing step comprises representing the AR document as an N-dimensional vector using a semantic vector space model (SVSM), and including: a step of creating a 2-dimensional (M×N) matrix containing M concepts for each document; a step of assigning weights to each concept based on a TF-IDF (term frequency-inverse document frequency) score representing the relationship between concepts corresponding to each pair of the 2-dimensional matrix; and a step of generating an SVSM for the document by applying the maximum weight to the superclass of the seed concept. Claim 8 In claim 5, the clustering step comprises a semantic information retrieval method including the step of clustering the AR document vectors using k-means clustering and displaying the information retrieval results as clusters. Claim 9 A semantic information retrieval method according to claim 1, wherein the step of performing the above further includes a step of preprocessing to convert a query and an AR document into a word sequence prior to the step of expressing the above. Claim 10 A semantic information retrieval method according to claim 9, wherein the preprocessing step is characterized by performing preprocessing through at least one of the following: a process of performing tokenization in which a query and an AR document are broken down into sentences and then broken down into individual words or terms, or a process of removing stop-words included in the query and an AR document. Claim 11 A computer program stored on a computer-readable recording medium to execute a method for retrieving semantic information of any one of paragraphs 1 to 5 and 7 to 10 on the computer device. Claim 12 A computer device comprising at least one processor implemented to execute a readable command on a computer device, wherein the at least one processor performs semantic information retrieval in an AR domain using an AR ontology composed of an augmented reality concept, and wherein the at least one processor represents an AR document as a vector using a semantic vector space model (SVSM) according to the concept of the AR ontology. Claim 13 A computer device according to claim 12, characterized in that the AR ontology includes hardware, software, tracking, interaction, and interface related to AR. Claim 14 A computer device according to claim 12, wherein at least one processor clusters AR documents into AR-related topics or concepts using the AR ontology. Claim 15 A computer device according to claim 12, wherein at least one processor clusters AR document vectors into the AR domain using k-means clustering.

Images