26

In the past, I've used apps that have suggested films, books, or music to me based on a set of items I've enjoyed that I provide to the app.

Is there anything like that for anime? Basically I'm looking for something in which I can input that I like anime X, Y and Z and for it to suggest that I might like anime A because it has similar themes, or because other people who like X, Y and Z have also said they like A.

Maroon
  • 14,564
  • 14
  • 71
  • 137
Omar Kooheji
  • 535
  • 2
  • 5
  • 9

8 Answers8

30

Anime-Planet and MyAnimeList have user submitted suggestions for shows on a "if you liked X, you might like Y" setup.

As a developer myself, this is something I've wanted to look into making, and MyAnimeList seems to be the most common site to be used for an engine as they provide an API to programmers.

Anime-Planet Recommendations:

Per-show Recommendations: enter image description here

Update: Anime-Planet now have version 1 of a personalized recommendation engine that works off what you have already seen

enter image description here

MyAnimeList Recommendations:

Per show recommendations

enter image description here

There are a couple of engines based off of MyAnimeList that people have made such as AnimeRecs (which is down at the moment) or mal.oko.im

There's also a new anime cataloguer Hummingbird which gives you a range of recommendations based on what you've watched, what genres you like, which are your favourites. It seems to be the easiest to access and isn't soley show-specific.

Hummingbird Suggestions:

enter image description here

Other Sources Other websites which provide anime have recommendations engine built in, Usually these are based off of what similarly interested people have enjoyed also.

Crunchyroll (Thanks @David_Starkey ):

Crunchyroll

Amazon:

enter image description here

Toshinou Kyouko
  • 34,175
  • 26
  • 160
  • 297
  • 1
    [crunchyroll](http://www.crunchyroll.com/) also has a "Viewers Also Liked" section on the anime pages. – David Starkey Feb 06 '14 at 19:14
  • MAL has personalised recommendations now on the front page, if you're logged in, although I don't know if one's anime/manga lists have to reach a certain length first before recommendations can be properly generated. FWIW I'm happy with the ones I got, insofar as they're not _unreasonable_ as recommendations. – Maroon Mar 26 '17 at 16:09
6

Just to add to the list, there's also anidb which will also provide related, similar, and random anime recommendations

Arty-chan
  • 410
  • 3
  • 9
4

graph.anime.plus is a great analyser for your myanimelist profile.
On top of giving you a great deal of information on your habits and breakdowns on what you've listed it appears to just trawls through your favs and pulls the mal recommendations (it'd be damn cool if itI've found a damn cool tool that finds similar-minded people and recommend things you haven't seen that they liked).

Another thing it does in the same tab is show you related items that you haven't seen yet, which may yield a sequel or cool OVA you missed



AniPlan does both of these too

2

I love this site: http://animesuggestions.com/

You just type any anime you like, e.g. "One Piece" or "Hunter x Hunter", and it shows you a bunch of similar animes from which you can choose whatever catches your eyes, I guess.

(I only tried it once and I searched "Nagi no Asukara" and found "Elfen Lied", and I really liked it, okay?)

Gao
  • 9,603
  • 8
  • 57
  • 96
Name
  • 21
  • 1
1

There aren't really dedicated "suggestion engines" for anime. your best bet is looking through places like [MyAnimeList] or [AnimePlanet]. Otherwise, you can go through something like this, scroll through to something that has genres you like, and look for whether or not it is highly rated.

In the end, there are plenty of websites you can go to that will give you info about series, recommendations, suggestions, etc., but nothing with a dedicated system. Just use a bunch'a different sites, and you won't go wrong.

DubFanatic
  • 11
  • 3
  • 1
    There are a bunch of recommendation sites (e.g. Hummingbird iirc, Tastekid) that may also involve anime recommendations. – Maroon Aug 28 '14 at 02:09
1

adding another answer since this is a drastically different way to go about it

You can use affinity.animesos.net to search for other people that share your tastes (based on mal's affinity).

enter image description here

Wow, someone with >95% affinity! Click their name to see their profile, then simply click the "~ shared" link.

enter image description here

I recommend sorting by score and first looking at what you two agreed on and what you didn't to be able to gauge where their tastes will apply to you, then just scroll down to the "Unique to [them]" section to add some of their favourites to your plan-to-watch!


Disclosure though; see how the affinity search said we only shared 11, yet mal says we share 79? Mal includes the unscored entries as items able to be common between people.

Here's a link to the comparison page, we might not be a great pair to extrapolate our opinions to each other.
So I bumped the search requirements to only use people with 50 shared titles, and that dropped my closest match to 69% affinity, but have a look at our comparisons page.

This should be a much more useful match to find new anime, especially if I work my way down the result list and see any titles that continue to pop up in a lot of peoples lists. Good luck!

1

Yes there are engine's such as that. A good sample would if you are a owner of a filled in Myanimelist. If you have that can go to mal graph mal.oku.im and see recommended. I attached my own recommended as a sample for this. http://mal.oko.im/d3ullist/recs,anime

Then again in the end it all depends on your own taste´s.

Dimitri mx
  • 22,062
  • 32
  • 118
  • 200
  • 1
    As a head's up, it looks like [the site is closing down within the next two months...](http://myanimelist.net/forum/?topicid=1373476) – Maroon Apr 17 '15 at 16:56
0

AniDB also has a recommendation feature. To use it, you need to add the series and episodes you watched to your AniDB MyList. Most importantly, you also need to rate the series from 1 to 10 (10 being best).

The system then uses some hidden algorithm based on what you voted and what others voted and thereby predicts your votes for series you have not yet watched. You can even remove recommendations from the prediction list if you say ‘I’m not going to watch that, stop telling me I’m going to rate it with a 9’.

Note that for recommendations to work best, you should attempt to vote in a way that puts your median at 5.5 and resembles a standard distribution. While the system ‘corrects’ votes for overall series ratings, if you use exaggerated votes you will also receive exaggerated vote predictions (such as 13, which is obviously larger than 10).

Jan
  • 966
  • 1
  • 10
  • 31