‘Customized content’ recommendation…Paid broadcasting, OTT curation service in match
Kim Ji Young | kjy@ | 2017-10-27 11:16:32
The paid TV industry is introducing a `curation` service which recommends content that matches the viewer`s taste. It is a strategy to target consumers who prefer personalized content rather than simple organization in a one-person media environment.
According to industry sources on October 26, major service providers who are engaged in online video (OTT) service provide content recommendation curation service that analyzes users` viewing patterns. KT Skylife combined WATCHA PLAY`s content recommendation engine with `TV`, an OTT service for TV. CJ HelloVision`s "Visualization", which will be released next month, uses Big Data and AI based machine learning. It is a strategy of transforming into an intelligent TV which `understands personal taste and recommends content` as you use it.
Mobile carriers operating a mobile OTT service also have a curation function. KT`s `ole tv mobile` has recently been providing services to recommend video-on-demand (VOD) services based on the customer`s viewing history. SK Broadband `s` corn` has different screens based on age, gender, and preferences of users.
The curation service recommends information or goods considering demographic factors such as gender, age, interests, and mood. In recent years, the scope of providing news, music, video, and other contents has been expanded. It collects, organizes, and edits contents according to the viewpoint or tendency of the individual among many contents to provide content that the user prefers. In the situation where contents are overflowing, it plays a role of enhancing the user convenience and the possibility of VOD purchase. Because of this, curation services are becoming more sophisticated by combining new technologies such as AI and O2O beyond simple information accumulation.
The content curation service was launched by overseas operators. NeXTrix has an evolutionary algorithm of a deep-running approach that gets closer to the user`s taste as the content is watched. It is a method of systematically classifying over 30,000 contents into genres, characters, and stories and analyzing data such as user`s viewing time, behavior, and preferred contents to find intersections. Beyond finding your favorite content, it even filters out unwanted information. 70% of Netflix users are selecting the following content as a recommended video. In the case of Netflix, the industry estimates that sales by curation are worth about $ 1 billion (about 1.14 trillion won).
Comcast, on the other hand, the largest cable operator in the United States, uses technologies such as AI, deep running, big data, and cloud to recommend programs and image classification. Comcast`s Deep Learning software provides customized menus to recommend TV shows and movies based on your past watch history. At the GPU Technology Conference in May, Jan Neumann, Comcast Technical R & D Director, said, "GPU-based deep-running using natural language processing, image recognition and video analysis driven by GPU- I can find episodes of TV series that I have not been able to do and can quickly summarize the highlights of TV programs and sports events. "
The Korean content curation service is still in its infancy. The more detail data the user has, the higher the recommendation accuracy. Domestic services tend to depend on one-dimensional information such as gender, age, and preference genre. An industry representative said, "Overseas, various algorithms such as clustering are being developed to compare with similar users even when there is insufficient data beyond the step of collecting data. Korean curation services are still in the process of developing algorithms , Real-time data analysis, and intelligent personalized services. "
By Kim Ji Young kjy@
[ copyright ⓒ The Digitaltimes ]
According to industry sources on October 26, major service providers who are engaged in online video (OTT) service provide content recommendation curation service that analyzes users` viewing patterns. KT Skylife combined WATCHA PLAY`s content recommendation engine with `TV`, an OTT service for TV. CJ HelloVision`s "Visualization", which will be released next month, uses Big Data and AI based machine learning. It is a strategy of transforming into an intelligent TV which `understands personal taste and recommends content` as you use it.
Mobile carriers operating a mobile OTT service also have a curation function. KT`s `ole tv mobile` has recently been providing services to recommend video-on-demand (VOD) services based on the customer`s viewing history. SK Broadband `s` corn` has different screens based on age, gender, and preferences of users.
The curation service recommends information or goods considering demographic factors such as gender, age, interests, and mood. In recent years, the scope of providing news, music, video, and other contents has been expanded. It collects, organizes, and edits contents according to the viewpoint or tendency of the individual among many contents to provide content that the user prefers. In the situation where contents are overflowing, it plays a role of enhancing the user convenience and the possibility of VOD purchase. Because of this, curation services are becoming more sophisticated by combining new technologies such as AI and O2O beyond simple information accumulation.
The content curation service was launched by overseas operators. NeXTrix has an evolutionary algorithm of a deep-running approach that gets closer to the user`s taste as the content is watched. It is a method of systematically classifying over 30,000 contents into genres, characters, and stories and analyzing data such as user`s viewing time, behavior, and preferred contents to find intersections. Beyond finding your favorite content, it even filters out unwanted information. 70% of Netflix users are selecting the following content as a recommended video. In the case of Netflix, the industry estimates that sales by curation are worth about $ 1 billion (about 1.14 trillion won).
Comcast, on the other hand, the largest cable operator in the United States, uses technologies such as AI, deep running, big data, and cloud to recommend programs and image classification. Comcast`s Deep Learning software provides customized menus to recommend TV shows and movies based on your past watch history. At the GPU Technology Conference in May, Jan Neumann, Comcast Technical R & D Director, said, "GPU-based deep-running using natural language processing, image recognition and video analysis driven by GPU- I can find episodes of TV series that I have not been able to do and can quickly summarize the highlights of TV programs and sports events. "
The Korean content curation service is still in its infancy. The more detail data the user has, the higher the recommendation accuracy. Domestic services tend to depend on one-dimensional information such as gender, age, and preference genre. An industry representative said, "Overseas, various algorithms such as clustering are being developed to compare with similar users even when there is insufficient data beyond the step of collecting data. Korean curation services are still in the process of developing algorithms , Real-time data analysis, and intelligent personalized services. "
By Kim Ji Young kjy@