a anime rec chart

What are the similarities between Rakudai and Banished from Hero's Party?

- Rakudai is about magic warfare VS Banished from hero's party is about medicinal remedies/drugs

What are the similarities between Kimi No Na Wa and Byousoku 5 centimeter?

Both directed by the brilliant mind of director Makoto Shinkai, Kimi No Na Wa and Byousoku 5 Centimeter share a lot of similarities. For some odd reason, both films feature beautiful space scenes with comets. Both films are about star-crossed lovers wanting to meet each other. The main characters have a strong bond, even when they live far away from each other. The endings of both films are beautiful. I definitely recommend Byousoku 5 Centimeter if you love Kimi No Na Wa

Where does Katanagatari take place?

Albeit Katanagatari is more on the fictional side whereas Heike monogatari is historic, both stories take place in feudal Japan wherein we are allowed to take a peek into the daily lives of its people, their beliefs and cultures, traditional cities and architecture along with landscapes tainted by battle not discriminating between large or small scale. The cycle of life and death is quite prevalent. Keeping sacred treasures and the family head safe for the sake of honor compared to the tale of seeking out ancient sword relics across Japan for fame and glory will leave the viewer contemplate the fickle mischiefs brought upon by karma once the unstoppable dices of fate are cast.

How many times did I cry watching K-On?

Philosophical works, 2000s cults but on different levels, while texhnolyze is shallow and poorly written k-on is magnificent and deep, I cried 9181 times watching.

Why is Hiiragi Seiichi bullied?

High school student Hiiragi Seiichi is bullied by his classmates for being a "loser." One day, his entire school is suddenly transported to a video game-like world of swords and sorcery. When he accidentally eats "the Fruit of Evolution," his life as a successful "winner" begins.

Who is Yoh Asakura?

On a starry night, middle-schooler Manta meets Yoh Asakura, a young shaman who has the ability to talk to spirits , and also has the ability to let them possess him to use their powers. Yoh has a heavy duty on his shoulders: he must become the Shaman King by winning the Shaman Fight, a tournament between shamans that happens only once every 500 years. Joined by his spirit medium fiancée Anna Kyouyama, his personal samurai spirit Amidamaru, and his other friends, Yoh must also face a larger evil: Hao, a powerful man who aims to create a new world with only shamans by becoming the Shaman King himself.

How old is Nagumi Koushi?

Nagumi Koushi is a 12-year-old boy who was abandoned by his father after their house caught on fire, leaving him to roam the streets penniless. Then one day, a girl named Minerva who finds him lying on the sidewalk, takes him to a women's college dormitory, one known for housing troublesome residents, and asks him to be their "Dormitory Mother." Surrounded by older women, forced to deal with their idiosyncrasies, Koushi starts his new (somewhat ecchi) life!

What does an orb do?

In the beginning, an "orb" is cast unto Earth. "It" can do two things: change into the form of the thing that stimulates "it"; and come back to life after death. "It" morphs from orb to rock, then to wolf, and finally to boy, but roams about like a newborn who knows nothing. As a boy, "it" becomes Fushi.

Data (Continued)

It's important to note that all of this data is explicit user feedback from a site that requires users to register in order to rate shows.

Architecture

As a brief explanation of the overall architecture, Once the Score and Review scrapers have been used to generate full data-lake of user scores and Anime Reviews, there is multi-step process used to generate a recommendation. To show the larger architecture before going into the components:

Generating Possible Recommendations

Anime_Score_Sim shows the code for this. First, we drop all 0's. Next, we find the average score for each user, and subtract that score from every user's score. This is normalize the data (Without doing this, user's having different ideas of what constitutes a 6 vs a 7 vs 8, adds a lot of noise to the data).

Predicting User Scores

There are 3 pre-trained neural nets. Each net has been trained to predict one type of score: Score, Usr Scaled Score, Anime Scaled score. The Nets are loaded with their trained weights, only the user embedding layer is unlocked . Each neural net predicts the user’s score for every possible rec (from the item-item methods).