Please enable JavaScript.
Coggle requires JavaScript to display documents.
Idea Generation, We start with the static data: an accurate and detailed…
-
We start with the static data: an accurate and detailed description of the game, The publisher/developer, and the genres
Next we add dynamic data: updates, events, tournaments, and everything the game developers have done to change the game since launch.
Finally, we will add in a touch of subjective data. This provides the LLM with information on how the game has been received by the audience
(professional reviews), and which activities are most preferred (youtube data)
Youtube Account Data
Channel Profile
The Channel Profile contains all of the data from a creator's channel, including the data from each of their videos, and information about their channel such as the number of subscribers, and the channel description.
Brand Insights Profile
The Brand Profile is designed to combine information from the Creator and Channel profiles in a meaningful way.
-
-
Steam Account Data
Gaming Profile
The Gaming Profile provides data on the game preferences of the creator. Attributes include genres, games, price range, skill level
Gaming Insights Profile
The Content Profile is designed to combine information from the Gaming and Channel profiles in a meaningful way.
For YouTube gaming creators, this means determining which games are found in which videos.
This profile will begin by determining which games are found in a specific creator's videos. Once we know the game found in the video, we can determine which gameplay concept, or combination of concepts are present in the video.
This provides us with information on which games are receiving the best feedback for a given content creator. This provides us with similar information on which Gameplay concepts are receiving the best feedback for a given creator.
Finally, this profile will summarize how the gaming data has impacted the engagement ratings for a given creator's videos.
With an understanding of which games and concepts are present in each video, we can analyze the engagement metrics (likes, views, sentiment analysis of comments) to determine which games, genres, and even specific concepts are best for a given target audience.
-
Bird Account Data
Creator Profile
The Creator Profile contains all of the data about the true personality of the creator off-screen. Attributes include the OCEAN personality assessment, and the personality goals of the creator.
-
To do this, we must decide which games to research each day, and how to research them.
We will focus on one list of general, popular games and a second list for niche games that our customers prefer.
For each game in the list, we will perform research to gain a deeper understand of the game.
This should cover all of the desired data for the purpose of understanding the logic found in each game. This includes the activities, mechanics, and events among other logical information.
We start by gathering a list of the 50 most played and most bought games on Steam. The most popular games are the ones viewers look for from content creators.
The first list is missing smaller games which our customers may prefer, but there are too many of these. So, we will gather the 50 most played / highest reviewed games from our customer's Steam Library
Currently, we just give all of this data to chatgpt with a prompt. How will it translate to an AI model?
To start, we must figure out which games this user enjoys, and what kind of personality they have.
-
After finding their favorite games, the next step is to gain an understanding of their personality
We will use their Steam Library to find their favorite games (simple, no AI work)
Finally with a list of game concepts from their favorite games, and an understanding of their personality, we can generate personalized ideas.
This is currently done with the data described and a prompt given to chatgpt. It will eventually work with our custom model.
-
-
-
-
-