AI Application Cases


    AI consultation system building cases

    Counselor and Q&A service for customers – NH Bank

    In the financial sector, a variety of customer service channels, including mobile services, are expanding, and the demand for services corresponding to the latest AI-based technology trends is increasing. It is said that NH Nonghyup Bank should increase the number of AI counseling service channels. Nonghyup Bank Customer Happiness Center aims to build a system infrastructure to expand service channels in order to adapt to all the environments that require AI counseling services for counselors and customers, both online and offline. It is expected that the Customer Happiness Center will play a pivotal role in expanding artificial intelligence services, thereby setting up an example in the entire financial sector.


    < NH Bank AI consultation system summary >

    ① Business contents
    1. Providing Q&A service for customer consultation knowledge. The service receives a question from the customer through the mobile app chat service, finds the corresponding answer, provides it in the form of text, or connect to a service such as account inquiry or transfer: customer consultation Q&A service building.
    2. It provides an answer to a question by converting counseling content into text in real time. Unlike the cases in which the counselor directly asks a question, it is possible to quickly respond to customer needs by referring to the feedback by the AI ​​system during the consultation: building a real-time telephone consultation support service.
    3. The AI ​​system first receives the customer’s call and provides a simple answer or connects to a professional counselor. Interactive counseling is provided through a dialogue model, and usual questions are answered by the Q&A system: building AI Virtual Counseling Service ‘Callbot’.
    4. It provides customer consultation and guidance services in which a guide robot connects with APIs provided by Nonghyup Bank AI system. Providing customer consultation by setting up a dialogue model for consultation contents and connecting with the Q&A system: guide Robot Consultation Service.

    < Real-time telephone consultation and customer consultation service >


    < Virtual consultation Callbot service and robot consultation service >

    ② Application technology and solutions
    1. AI Suite product application
      Consultation services include AI Suite’s natural language understanding engine, deep QA engine, dialogue processing engine, and machine reading comprehension engine to process conversation flow and provide Q&A for service scenario.
    ③ Main accomplishments
    1. Building a system infrastructure in which services are expanded across all channels to employees and customers through the AI counseling system
    2. Interactive consultation service is provided through deep QA engine and dialogue processing engine. The conversation allows the system to identify and answer customer intentions, improve the ability to answer simple questions that are not clear, and continue to consult with additional questions and provide smooth services.
    3. Enhancement of counseling quality and specialization of counselor’s duties by providing counseling assistants
    4. The consultation talk service can be introduced in the Nonghyup Bank mobile apps, All-One Bank and Smart Banking to expand the service to the users.
    5. Korea’s first artificial intelligence telephone call service Call Bot was developed and commercialized. It is possible to achieve a faster response than the existing ARS guidance, while the counselor can concentrate on professional counseling. It was confirmed that the pilot robot system could provide interactive counseling and knowledge services through the operation of the guide robot in offline channels such as branches.

    Intelligent knowledge management systems building examples

    Intelligent knowledge management system – NH Bank

    The knowledge management system, which manages and shares the work knowledge required for customer consultation at customer call centers and branches, has been improved, so that it can be used in AI systems. The ease with which knowledge is generated, managed, searched, etc. is enhanced, and when knowledge content is generated, it is specifically managed according to the work field and attributes. So it is converted and stored as Knowledge-Graph-based knowledge base. This automatically converts natural language content into machine readable knowledge data. This process is based on AI-based automatic knowledge extraction, knowledge data extraction, semantic complex inference, and then it is converted and stored in the knowledge base.


    < Intelligent knowledge management system summary >

    ① Business contents
    1. In consideration of knowledge extraction and conversion into knowledge graph, work categories are defined in a hierarchical structure, and a detailed content management tool is built to manage and create knowledge contents for each category: Knowledge Content Creation Management.
    2. It automatically extracts knowledge data candidates by analyzing knowledge content written in unstructured text. This is an automatic knowledge extraction process in which the target knowledge is identified through natural language processing and data is extracted through deep learning-based learning models. The extracted knowledge data is converted into a triple form so that it can be stored in the knowledge graph: Automatically Extract Knowledge Data.
    3.  Building a quality management environment in which the automatically extracted knowledge data can be verified and edited. Constant improvement of quality and performance in automatic knowledge extraction: Knowledge learning verification and management.

    < Automatic knowledge extraction and management through the intelligent knowledge system >

    ② Application technology and solutions
    1. AI Suite product application
      In order to extract and manage knowledge from knowledge content, AI Suite’s natural language understanding engine, knowledge learning engine, and complex inference engine are applied.
    ③ Main accomplishments
    1. Automatic knowledge extraction and knowledge extraction are technical fields that still require a lot of efforts in research and development. Various methods including language processing and machine learning were studied to improve the quality, and as for the extraction of necessary knowledge from the contents of the domain, results were satisfactory. It could be regarded as a successful example of applying automatic knowledge extraction to real services.
    2. Since the knowledge content managed in the knowledge management system can be utilized in the artificial intelligence system, the knowledge utilization becomes high. In addition, unifying the knowledge information that has been managed separately reduced the cost of knowledge maintenance and made it easier to use.

    Chatbot service building examples

    Chatbot service introducing project for internal business support – KEPCO KDN

    KEPCO KDN introduced Chatbot service, which applied basic dialogue processing engines such as high-performance natural language processing, intent/entity mapping technology, and dialogue modeling. Efforts have been made to extend the system to provide effective Q & A for a variety of knowledge in the future, which includes linking knowledge graphs. For the convenience of KEPCO’s employees, it was based on an easy-to-access messenger platform. The goal is to provide personalized answering services beyond simple knowledge in the fields of business trips, ICT equipment rentals, and connection of workers who need contact for work.


    < KEPCO KDN Chatbot service block diagram >

    ① Business contents
    1. Collecting frequently asked questions from employees and building a terminology dictionary: domain dictionary building.
    2. Building a dialogue model so that Chatbot can operate according to each work process: work process dialogue model building.
    3. Various intents derived from consultation are applied and sentences from the consultation are separately treated to increase the hit rate in intent extraction: sentence analysis.
    4. Building business trip expense guidance and ICT equipment rental work guidance: guidance service implementation.

    < KEPCO KDN Chatbot service >

    ② Application technology and solutions
    1. AI Suite’s dialogue processing engine and selective application of deep Q/A technology
    ③ Main accomplishments
    1. Improving work efficiency on main work by reducing human resources for simple support work with the help of knowledge-based work support
    2. Creating complex logical flow charts to take specific actions for different work situations.
    3. Not only Q&A on business procedures but also business application and reservation through groupware interworking are possible on one messenger platform.

    Recommendation service building examples

    Broadcast content recommendation system building – KBS

    Due to the open technology environment, service environment such as broadcasting/communication integration, and the evolved content consumption structure, there is an increasing need for an environment in which necessary content can be easily found and consumed. That’s why we developed a customized content recommendation technology based on the user’s characteristics and situations. In this example, we explored and developed user-tailored broadcast content recommendation search for broadcast video contents and related contents using language analysis, search, mining, and semantic technologies.
    Broadcast content recommendation search system is composed of four systems, data collection, content search, content recommendation, API server and they can be divided into independent systems. Independent systems are divided into data collection, content indexing, content analysis, content navigation, content search, user analysis, and content recommendation.


    < Broadcast content recommendation system block diagram >

    ① Business contents
    1. Collection and storage management of broadcast metadata and related information at KBS for content analysis, user profile information for broadcast analysis, broadcast service history, and user feedback information: content and preference data platform.
    2. By analyzing the content consumption history data of the user, the degree of similarity in patterns between two users who consumed different kinds of content is derived and the subsequent content recommended search result is provided: Recommended Search Platform.
    3. Two algorithms are implemented. One is a content-based algorithm that recommends content based on a user’s input (search term, menu, content) by analyzing similarity or association between the contents based on the characteristics of the collected content metadata. The other is user based recommendation algorithm using user-based collaborative filtering technique and item-based collaborative filtering technique based on user’s personal preference information and profile: Recommended search algorithm development.
    4. Implementing data and service management features required for recommendation systems such as collection management, index management, project management, API management, and content management: Recommended Search Management.
    5. Development of recommendation search interface and implementation of utilization service such as program related menu and related recommendation page, content/keyword/person- based recommendation search, customized recommendation work, and recent recommendation work: Recommended Search Utilization Service.

    < Broadcast content recommendation service screen >

    ② Application technology and solutions
    1. AI Suite’s natural language understanding engine and complex reasoning engine
    ③ Main accomplishments
    1. Identifying trends in research for recommendation search for broadcasting content recommendation services, and planning, studying and developing components of recommendation systems, recommendation algorithms, target features and services based on existing recommendation system examples.
    2. Collecting broadcast content data, metadata and user history information based on modeling according to broadcast metadata standards. User history information implements a recommendation model according to each recommendation algorithm, and broadcast content data and metadata implement a recommendation model using text mining and statistical techniques.
    3. Content recommendation search is applied to match OHTV recommendation service, and it is applied to recommendation search web page to build utilization service of recommendation search system.
    4. To study how to use the recommendation search engine, it is applied to KBS integrated CMS. We developed a user-based recommendation search service that provides five recommendation features: consumption history-based recommendation, preferred person recommendation, preferred genre recommendation, group/gender/age recommendation, and composite recommendation.

    Artificial smart news recommendation app service building – Ziny News

    News and social content are the heart of big data that can be analyzed through artificial intelligence. Saltlux has built the Genie News service, a smart news app that analyzes more than 1,200 domestic and online news and social media in real time with the AI ​​engine, recommending and providing personalized information to 300,000 individual users.

    Genie News AI engine reads 7 million news and blogs a day, as people do, automatically classifies them into more than 500 categories, and automatically recognizes emerging issues, providing customized service. In particular, the personalized news feature uses Deep Learning technology to learn what each user reads, as well as registered keywords of interest, and automatically predicts and recommends the content that users may like based on what the engine learned. It is to create a kind of artificial persona for each user and a personalized neural network in which each individual gets content recommendation. The point is that the AI engine does personalized news recommendation service through an anonymous artificial neural network without collecting user’s personal information.


    < Ziny News service introduction screen >

    ① Business contents
    1. Deep learning-based user customized content recommendation
      Deep learning-based psychography learning and demography-based integration recommendation technology study and news recommendation service development based on it
    2. Intelligent curation of deep news and social content
      Extracting core content and major issues(themes) through large-scale content real-time analysis
      Classifying content by category and providing semantic technology-based related/similar content analysis
    3. Providing more advanced user experience(UX)
      Implementing UI/UX that enhances user convenience and content readability.
      Personalizing interest content and implementing social sharing features
    ② Application technology and solutions
    1. AI Suite’s natural language understanding engine and complex reasoning engine
    ③ Main accomplishments
    1. Providing in real time today’s issues that you shouldn’t miss. Confirming today’s hot issues by providing various styles of card news, such as eliminating repetition of the same news, ranking news, just-ins, 3-min briefings, etc: cover story.
    2. Collecting in real time news from major newspapers and providing various content including social media content and things that interest users Providing news and social media content by selecting only categories by subject/topic that interests users: news/social stream.

    < Ziny News content curation service >

    3. Based on the registered keywords of interest, the content by subject that the viewer might be interested in is selected and provided regardless of news or social media: Customized News.

    4.  Daily briefings with important information such as politics, economy, world, IT and sports: 3-min briefing.

    5. Individual users are analyzed in terms of preferences and interesting contents that they may prefer are provided. A variety of contents, including topics, ages and genders, are provided using recommendation algorithms: Ziny recommendation.


    < Ziny News content recommendation service >

    6. Provides an ad-free smart view considering the user’s convenience and readability: Article Viewer.

    7. Articles are analyzed by date and automatically categorized in chronological order so that they can be viewed at a glance. Providing a variety of things that the user might be interested in by recommending other related topic articles: News Chronicles / Related News.


    < User-friendly feature of Ziny News >

    Overseas examples

    [Japan] AI Consulting Systems

    Saltlux introduced AI counseling systems, virtual counseling assistant services to the Japanese financial and aviation industries, and commercialized AI technologies such as machine learning, knowledge reasoning, and deep Q/A. Thprofit, and increased customer satisfaction.


    < Japanese AI consultation system building examples >

    E introduction of an artificial intelligence counseling system prevented customer churn, allowed the system to solve problems on its own, improved customer satisfaction, enhanced corporate image, increased operating

    ① Business contents
    1. We built a service with a concept of concierge in which, just like a real counselor, the system understands the intentions of people’s questions, finds appropriate answers, and provides those answers. To deal with Mizuho Bank customers’ problem solving, Push-type navigator Mi-na service is provided on its business website: Mizuho Bank Virtual Counseling.
    2. Consultation services were added on the Monex Securities webi>: Monex Securities.
    3. Nipponkoa Insurance CompanyNipponkoacompany’s official website set up an ‘everyday helper’ service that helps deal with customers’ issues (日々乃まもり)site.’: Nipponkoa Insurance Co.
    4. Amy service was set up on ANA SKY WEB of All Nippon Airways (ANA, All Nippon Airways): All Nippon Airways Co., Ltd.

    < AI Suite overseas application example >

    ② Application technology and solutions
    1. AI Suite product application
      Application of natural language understanding engine and deep Q/A engine of AI Suite in consultation service
    ③ Main accomplishments
    1. Preventing customer churn
      Providing the customer, during the inquiry, with the necessary information to prevent the customer from leaving the service page.
    2. Decrease in the number of consultations
      After the concierge service began resolving the call, the number of consultation calls and inquiry mails decreased.
    3. Improvement of level of customer satisfaction
      Supplement the lack of knowledge and improving quality through service monitoring. The evaluation rate by service users continue to rise, leading to increased actual customer satisfaction.