Most Frequently Similar on Lesbian Sex stories

On a similar note, there's a big (in views/popularty) mom/son story posted today, with the Lap theme. Those always get tons of views.

A tag they used was 'party'

.

I note with approval that the author followed the SimonDoom "You can't go wrong by being completely obvious" theory of story-titling. The title is "Mom Son Lap Halloween." It has over 25,000 views less than 12 hours after publication.
 
A brother/sister lap story was posted a few days ago with a similar view range.

The lap story always seems to work.
 
As this thread has shown, the question of what stories end up on a Similar Stories list really has two aspects. First there is the technical question: how does the website go about choosing which stories to put on the list? Second there is the popularity-contest question: why do some stories show up on Similar Stories lists more often than others?

I've been hacking away at the technical question, and I think I've figured it out. Consider this story:

Story 5. Riddle of the Copper Coin, by Bramblethorn, Sci-Fi & Fantasy, 03/09/2016. This story has been favorited by 48 people. Its Similar Stories list is as follows:

[tr][td].[/td][td].[/td][td]
Fans in​
[/td] [td]
Date​
[/td][/tr] [tr][td]____[/td][td]Similar Story_____________[/td] [td]Common[/td] [td]__Published[/td] [td]__Cat[/td] [td]__Hearts[/td] [td]_MFList[/td][/tr] [tr][td]
1.​
[/td] [td]A Proper Send-off[/td] [td]
18​
[/td] [td]
11/12/11​
[/td][td]
LS​
[/td] [td]
2007​
[/td] [td]
2​
[/td][/tr] [tr][td]
2.​
[/td] [td]Lovers Without Realizing It[/td] [td]
17​
[/td] [td]
03/15/09​
[/td][td]
LS​
[/td] [td]
2486​
[/td] [td]
1​
[/td][/tr] [tr][td]
3.​
[/td] [td]A Ghost of a Chance[/td] [td]
14​
[/td] [td]
11/16/17​
[/td][td]
LS​
[/td] [td]
414​
[/td] [td]
44​
[/td][/tr] [tr][td]
4.​
[/td] [td]The Case of the Sneaky Valentine[/td] [td]
14​
[/td] [td]
01/24/15​
[/td][td]
LS​
[/td] [td]
871​
[/td] [td]
4​
[/td][/tr] [tr][td]
5.​
[/td] [td]A Benign Something[/td] [td]
13​
[/td] [td]
10/23/05​
[/td][td]
LS​
[/td] [td]
1006​
[/td] [td]
14​
[/td][/tr] [tr][td]
6.​
[/td] [td]Could You Be Mine?[/td] [td]
12​
[/td] [td]
04/06/18​
[/td][td]
LS​
[/td] [td]
558​
[/td] [td]
5​
[/td][/tr] [tr][td]
7.​
[/td] [td]Like A Secret Being Told[/td] [td]
12​
[/td] [td]
06/23/10​
[/td][td]
LS​
[/td] [td]
1050​
[/td] [td]
6​
[/td][/tr] [tr][td]
8.​
[/td] [td]When You're Near Me[/td] [td]
12​
[/td] [td]
08/09/06​
[/td][td]
LS​
[/td] [td]
899​
[/td] [td]
16​
[/td][/tr] [tr][td]
9.​
[/td] [td]Double Blind Date[/td] [td]
11​
[/td] [td]
08/14/14​
[/td][td]
LS​
[/td] [td]
713​
[/td] [td]
10​
[/td][/tr] [tr][td]
10.​
[/td] [td]Evergreen Kiss[/td] [td]
11​
[/td] [td]
01/24/14​
[/td][td]
LS​
[/td] [td]
437​
[/td] [td]
43​
[/td][/tr] [tr][td]
11.​
[/td] [td]A Broken Woman Can Still Heal[/td] [td]
11​
[/td] [td]
11/14/13​
[/td][td]
LS​
[/td] [td]
337​
[/td] [td]
-​
[/td][/tr] [tr][td]
12.​
[/td] [td]Mary's Innocent Passion[/td] [td]
10​
[/td] [td]
11/26/18​
[/td][td]
LS​
[/td] [td]
232​
[/td] [td]
-​
[/td][/tr] [tr][td]
13.​
[/td] [td]Hope a Little Longer[/td] [td]
10​
[/td] [td]
04/06/18​
[/td][td]
LS​
[/td] [td]
231​
[/td] [td]
-​
[/td][/tr] [tr][td]
14.​
[/td] [td]Kisses from Hell[/td] [td]
10​
[/td] [td]
11/17/17​
[/td][td]
NH​
[/td] [td]
719​
[/td] [td]
-​
[/td][/tr] [tr][td]
15.​
[/td] [td]Satyr Play[/td] [td]
10​
[/td] [td]
03/27/17​
[/td][td]
SF​
[/td] [td]
737​
[/td] [td]
-​
[/td][/tr] [tr][td]
16.​
[/td] [td]Sabine[/td] [td]
10​
[/td] [td]
02/19/16​
[/td][td]
LS​
[/td] [td]
262​
[/td] [td]
-​
[/td][/tr] [tr][td]
17.​
[/td] [td]Natalie's Valentine[/td] [td]
10​
[/td] [td]
01/20/16​
[/td][td]
LS​
[/td] [td]
326​
[/td] [td]
-​
[/td][/tr] [tr][td]
18.​
[/td] [td]My One True Christmas Wish[/td] [td]
10​
[/td] [td]
12/07/15​
[/td][td]
LS​
[/td] [td]
554​
[/td] [td]
37​
[/td][/tr] [tr][td]
19.​
[/td] [td]The Cotillion[/td] [td]
10​
[/td] [td]
10/01/15​
[/td][td]
R​
[/td] [td]
838​
[/td] [td]
-​
[/td][/tr] [tr][td]
20.​
[/td] [td]Purely Sinful[/td] [td]
10​
[/td] [td]
10/07/14​
[/td][td]
EH​
[/td] [td]
1927​
[/td] [td]
-​
[/td][/tr]

[tr][td]Fans in Common = no. of people who favorited both Similar Story and Story 5[/td][/tr] [tr][td]Hearts = no. of people who favorited Similar Story[/td][/tr] [tr][td]MFList = position of Similar Story on 8Letters' Most Frequently Similar list[/td][/tr]

As you can see, the stories are listed in decreasing order of the number of fans they have in common with Story 5. The top story on the list has been favorited by 18 of the people who favorited Story 5, the second by 17, the next two by 14, and so on. Within each group, the stories are listed in reverse chronological order.

Category doesn't matter (although see caveat 3 below). Story 5 is in Sci-Fi/Fantasy, but only one of the Similar Stories is in that category. Popularity doesn't matter. Some of the stories near the top of the list have fewer hearts than some of the ones lower down. 11 of the stories are on 8Letters Most Frequent Lesbian Similar Stories List, but not all of them are, and, for example, the third most frequently Similar Story ("Bedding the Babysitter") is not on the SS list.

I've examined a number of Similar Stories lists, and they've all been organized this same way: ordered first by number of fans in common and second by publication date.

Reverse Engineering. The first thought that comes to mind is that the site finds all the stories that have fans in common with a given story and chooses the top ones for its SS list. Such an algorithm might go something like this:

- Assemble the combined Favorites List of all Story X's fans.
- Determine the distinct stories and count how many fans favorited each one.
- Edit the list if necessary according to the caveats below.
- Sort the resulting list by number of fans in common and publication date.
- Take the top 20 entries as Story X's Similar Stories list.

It's possible, however, that the site actually uses some other algorithm that finds some of the stories that have fans in common, but not necessarily all of them. In this case, for example, there might be other stories that the algorithm missed that have more fans in common with Story 5 than the ones listed in the table. Trying to figure out whether or not there are any missing stories turns out to be difficult, as discussed below. So I can't say for sure that this is the algorithm the site actually uses, but I'd bet it is.

Here are a few caveats to the above discussion.

Caveat 1. Everything here applies to the new-style SS lists (seen on the new (Beta) story reader). The old-style SS lists are not always identical to the new-style ones.

Caveat 2. The new-style SS lists are "series aware." Series are always referred to by their first chapter, and their ranking in the SS list seems to be based on the chapter with the largest number of fans in common

Caveat 3. The site does seem to do some filtering based on Categories. The one rule I'm pretty sure about is "I/T stories can't go on SS lists of non-I/T stories." This doesn't seem to be the case, though, for GM, LS, or Fetish. For example, I've seen a GM story on the SS list of a non-GM story. Whether or not there are other category filtering rules remains to be seen.

Caveat 4. The new-style SS lists are not updated immediately when a story is favorited. I have a story that was published in August, and its SS list seems to have only recently been updated from tag-based to favorites-based.

Caveat 5. If there are fewer than 20 favorites-based candidates, they will go into the top slots of the SS list, and the lower slots will be based on tags. As 8Letters and Bramblethorn have pointed out, sometimes there are not enough candidates of either type. In this case the SS list will be truncated.

Caveat 6. Some of the SS lists I've examined have slight discrepancies from the above rules that I can't currently explain.

Methodology. The number of fans in common between Story A and Story B can be determined by counting how many of the people who favorited Story A appear in the list of people who favorited Story B. The only way I know of to get the full list of people who favorited a story is to go to the last page of the story in the old (classic) story reader, have your browser show the page source code, and sift through the code for the list. I copied the lists into a spreadsheet to do the analysis for Story 5.

In order to verify that an SS list contains all the stories that have so many fans in common, you have to check every story on each fan's Favorites list. I've done this for some stories and generally found that their SS lists do contain all the stories most commonly favorited by their fans. This is not possible, however, if one of the fans is a "kazillion-favoriter." As Bramblethorn has pointed out, the Members Page truncates very long Favorites lists. As far as I can tell, the lists are truncated at 6000 entries, and they do not include the most recent ones. This makes it impossible to check the kazillion-favoriter's entire Favorites list.
 
I've been hacking away at the technical question, and I think I've figured it out. Consider this story:

Story 5. Riddle of the Copper Coin, by Bramblethorn, Sci-Fi & Fantasy, 03/09/2016. This story has been favorited by 48 people. Its Similar Stories list is as follows:
...

As you can see, the stories are listed in decreasing order of the number of fans they have in common with Story 5. The top story on the list has been favorited by 18 of the people who favorited Story 5, the second by 17, the next two by 14, and so on. Within each group, the stories are listed in reverse chronological order.

Looks like you've cracked it! The recommendations make sense for that story: it's posted in SF/F but also has strong elements of LS and Romance, and most of those similar stories look to be lesbian romances. Happy to provide a data point :)

Your proposed explanation seems particularly plausible since it's a fairly simple method, and one that would be computationally cheap to implement.
 
That is correct.

Edit: 6469 stories (over a third of the LS stories) have at least one silkstockingslover story as a Similar Story.

Is that true?
Seems crazy,
 
One of the intriguing questions about 8Letters' list of Most Frequently Similar Lesbian Sex Stories is how a story like "The Return of Dr. Mecuniam" ended up on it. "Dr. M" is not that highly rated (3.67 stars) and not that popular (5 hearts). How did it end up being the 34th most frequently cited Similar Story in the category? What is it about the story that led to its being selected as a fitting match for 372 other stories?

Well, it turns out that the only reason that "Dr. M" appears on so many SS lists is because it was one of the five last stories ever favorited by bigmike8204.

Big Mike was an avid pornophile who joined Lit on 8/30/2018 and proceeded to favorite stories by the boatload. He favorited so many, in fact, that he maxed out the Favorites list on his Member Page, and we're only able to see the first 6000 of them.

Big Mike's Favorites list actually seems to be more a reading list than a list of stories that he actually read. The part we can see consists of four swathes of stories, each arranged in alphabetical order. Stories 1-17 are miscellaneous lesbian stories, 18-1294 are female masturbation stories, 1295-2971 are lesbian incest stories, and 2972-6000 are general lesbian stories, only getting as far as the letter B. The most recent story on his Member/Favorites page is from 11/01/18.

He didn't stop favoriting at 6000, though. His name shows up on the Favorited-By lists of many stories that don't appear on his Member/Favorites page.

The most recent story that Big Mike favorited seems to be "The Wife, the Artichoke, and Jade," published on 4/21/19. He favorited stories up to that date, but as far as I can tell, none after it. For example, he favorited the early chapters of "The Newlins" up to chapter 13 (04/19/19), but he didn't favor chapter 14 (05/04/19) or any of the ones after that.

"The Return of Dr. Mecuniam" was published just a couple days earlier, on 04/18/19. It is apparently the fifth most recent story on Big Mike's full Favorites list. The way the Similar Story selection algorithm works, if a story has only one fan, its Similar Stories are taken from the five most recent stories on that fan's Favorites List. That means that any story that has Big Mike as its only fan will have "Dr. M" as a Similar Story. Because Big Mike favorited so widely, there are a lot of stories that have him as their only fan, 277 in LS in fact, according to 8Letters' post. These make up the lion's share of the stories that cite "Dr. M" in their SS lists. The others are presumably mostly stories whose fan lists include Big Mike and one or more of the other people who favorited "Dr. M."

There are actually four other stories from 8Letter's list that appear alongside "Dr. M" on many SS lists. They are "The Wife, the Artichoke, and Jade" (04/21/19), "What She Wants Is You" (04/20/19), "The Spring of 1984" (04/18/19), and "The Newlin's, Marcie & Mark" (Pt. 13, 04/19/19). These are presumably the other four most recent stories on Big Mike's full list. (An additional 15 stories appear on the extended SS lists.)

Basically, the reason that Big Mike's five most recent Favorites appear on so many SS lists is because he favorited so many less-popular stories where his Favorites list carries a lot of weight. The fact that "Dr. M" is included among those five is purely a consequence of the date he stopped favoriting. If he had stopped a week earlier or a week later, there would have been a completely different set of stories on all those SS lists.

Of course Big Mike favorited many higher-rated stories as well, but they usually had other fans whose Favorites lists outvoted his. That's why "Dr. M'' isn't often cited in the SS lists of higher-rated stories. And of course Big Mike isn't the only mega-favoriter. But somehow his mega-favoriting style seems to have had an oversized influence on LS Similar Story selection.

Has the presence of "The Return of Dr. Mecuniam" on so many SS lists driven a lot of traffic its way? Probably not. Comparing "Dr. M" (LS, 04/18/19, 3.76 stars, 5 hearts) to a comparable story that is not on 8Letters's list, "Limit of Sums" (LS, 05/18/19, 3.8 stars, 5 hearts), shows not much difference in the number of views (5.5k vs 4.9k). The stories that cite "Dr. M" are little read. There's not that much traffic to drive.

What does all this say about the way that Lit chooses Similar Stories? The basic idea of choosing Similar Stories based on favoriting clearly makes sense, and it probably works well a lot of the time. But when a story has only one fan, and that fan is a mega-favoriter? Take for example, "Annie's Girl Gets Her" (LS, 10/08/2000, 1 heart). It has "Dr. M" as a Similar Story, the only obvious similarity being that they both appear on Big Mike's list. Is "Dr. M" a better choice than one based on tags or an LS story chosen at random? Hard to say.
 
One of the intriguing questions about 8Letters' list of Most Frequently Similar Lesbian Sex Stories is how a story like "The Return of Dr. Mecuniam" ended up on it. "Dr. M" is not that highly rated (3.67 stars) and not that popular (5 hearts). How did it end up being the 34th most frequently cited Similar Story in the category? What is it about the story that led to its being selected as a fitting match for 372 other stories?
I like your explanation. It makes sense to me.
 
As this thread has shown, the question of what stories end up on a Similar Stories list really has two aspects. First there is the technical question: how does the website go about choosing which stories to put on the list? Second there is the popularity-contest question: why do some stories show up on Similar Stories lists more often than others?

I've been hacking away at the technical question, and I think I've figured it out. Consider this story:

Story 5. Riddle of the Copper Coin, by Bramblethorn, Sci-Fi & Fantasy, 03/09/2016. This story has been favorited by 48 people. Its Similar Stories list is as follows:

[tr][td].[/td][td].[/td][td]
Fans in​
[/td] [td]
Date​
[/td][/tr] [tr][td]____[/td][td]Similar Story_____________[/td] [td]Common[/td] [td]__Published[/td] [td]__Cat[/td] [td]__Hearts[/td] [td]_MFList[/td][/tr] [tr][td]
1.​
[/td] [td]A Proper Send-off[/td] [td]
18​
[/td] [td]
11/12/11​
[/td][td]
LS​
[/td] [td]
2007​
[/td] [td]
2​
[/td][/tr] [tr][td]
2.​
[/td] [td]Lovers Without Realizing It[/td] [td]
17​
[/td] [td]
03/15/09​
[/td][td]
LS​
[/td] [td]
2486​
[/td] [td]
1​
[/td][/tr] [tr][td]
3.​
[/td] [td]A Ghost of a Chance[/td] [td]
14​
[/td] [td]
11/16/17​
[/td][td]
LS​
[/td] [td]
414​
[/td] [td]
44​
[/td][/tr] [tr][td]
4.​
[/td] [td]The Case of the Sneaky Valentine[/td] [td]
14​
[/td] [td]
01/24/15​
[/td][td]
LS​
[/td] [td]
871​
[/td] [td]
4​
[/td][/tr] [tr][td]
5.​
[/td] [td]A Benign Something[/td] [td]
13​
[/td] [td]
10/23/05​
[/td][td]
LS​
[/td] [td]
1006​
[/td] [td]
14​
[/td][/tr] [tr][td]
6.​
[/td] [td]Could You Be Mine?[/td] [td]
12​
[/td] [td]
04/06/18​
[/td][td]
LS​
[/td] [td]
558​
[/td] [td]
5​
[/td][/tr] [tr][td]
7.​
[/td] [td]Like A Secret Being Told[/td] [td]
12​
[/td] [td]
06/23/10​
[/td][td]
LS​
[/td] [td]
1050​
[/td] [td]
6​
[/td][/tr] [tr][td]
8.​
[/td] [td]When You're Near Me[/td] [td]
12​
[/td] [td]
08/09/06​
[/td][td]
LS​
[/td] [td]
899​
[/td] [td]
16​
[/td][/tr] [tr][td]
9.​
[/td] [td]Double Blind Date[/td] [td]
11​
[/td] [td]
08/14/14​
[/td][td]
LS​
[/td] [td]
713​
[/td] [td]
10​
[/td][/tr] [tr][td]
10.​
[/td] [td]Evergreen Kiss[/td] [td]
11​
[/td] [td]
01/24/14​
[/td][td]
LS​
[/td] [td]
437​
[/td] [td]
43​
[/td][/tr] [tr][td]
11.​
[/td] [td]A Broken Woman Can Still Heal[/td] [td]
11​
[/td] [td]
11/14/13​
[/td][td]
LS​
[/td] [td]
337​
[/td] [td]
-​
[/td][/tr] [tr][td]
12.​
[/td] [td]Mary's Innocent Passion[/td] [td]
10​
[/td] [td]
11/26/18​
[/td][td]
LS​
[/td] [td]
232​
[/td] [td]
-​
[/td][/tr] [tr][td]
13.​
[/td] [td]Hope a Little Longer[/td] [td]
10​
[/td] [td]
04/06/18​
[/td][td]
LS​
[/td] [td]
231​
[/td] [td]
-​
[/td][/tr] [tr][td]
14.​
[/td] [td]Kisses from Hell[/td] [td]
10​
[/td] [td]
11/17/17​
[/td][td]
NH​
[/td] [td]
719​
[/td] [td]
-​
[/td][/tr] [tr][td]
15.​
[/td] [td]Satyr Play[/td] [td]
10​
[/td] [td]
03/27/17​
[/td][td]
SF​
[/td] [td]
737​
[/td] [td]
-​
[/td][/tr] [tr][td]
16.​
[/td] [td]Sabine[/td] [td]
10​
[/td] [td]
02/19/16​
[/td][td]
LS​
[/td] [td]
262​
[/td] [td]
-​
[/td][/tr] [tr][td]
17.​
[/td] [td]Natalie's Valentine[/td] [td]
10​
[/td] [td]
01/20/16​
[/td][td]
LS​
[/td] [td]
326​
[/td] [td]
-​
[/td][/tr] [tr][td]
18.​
[/td] [td]My One True Christmas Wish[/td] [td]
10​
[/td] [td]
12/07/15​
[/td][td]
LS​
[/td] [td]
554​
[/td] [td]
37​
[/td][/tr] [tr][td]
19.​
[/td] [td]The Cotillion[/td] [td]
10​
[/td] [td]
10/01/15​
[/td][td]
R​
[/td] [td]
838​
[/td] [td]
-​
[/td][/tr] [tr][td]
20.​
[/td] [td]Purely Sinful[/td] [td]
10​
[/td] [td]
10/07/14​
[/td][td]
EH​
[/td] [td]
1927​
[/td] [td]
-​
[/td][/tr]

[tr][td]Fans in Common = no. of people who favorited both Similar Story and Story 5[/td][/tr] [tr][td]Hearts = no. of people who favorited Similar Story[/td][/tr] [tr][td]MFList = position of Similar Story on 8Letters' Most Frequently Similar list[/td][/tr]

As you can see, the stories are listed in decreasing order of the number of fans they have in common with Story 5. The top story on the list has been favorited by 18 of the people who favorited Story 5, the second by 17, the next two by 14, and so on. Within each group, the stories are listed in reverse chronological order.

Category doesn't matter (although see caveat 3 below). Story 5 is in Sci-Fi/Fantasy, but only one of the Similar Stories is in that category. Popularity doesn't matter. Some of the stories near the top of the list have fewer hearts than some of the ones lower down. 11 of the stories are on 8Letters Most Frequent Lesbian Similar Stories List, but not all of them are, and, for example, the third most frequently Similar Story ("Bedding the Babysitter") is not on the SS list.

I've examined a number of Similar Stories lists, and they've all been organized this same way: ordered first by number of fans in common and second by publication date.

Reverse Engineering. The first thought that comes to mind is that the site finds all the stories that have fans in common with a given story and chooses the top ones for its SS list. Such an algorithm might go something like this:

- Assemble the combined Favorites List of all Story X's fans.
- Determine the distinct stories and count how many fans favorited each one.
- Edit the list if necessary according to the caveats below.
- Sort the resulting list by number of fans in common and publication date.
- Take the top 20 entries as Story X's Similar Stories list.

It's possible, however, that the site actually uses some other algorithm that finds some of the stories that have fans in common, but not necessarily all of them. In this case, for example, there might be other stories that the algorithm missed that have more fans in common with Story 5 than the ones listed in the table. Trying to figure out whether or not there are any missing stories turns out to be difficult, as discussed below. So I can't say for sure that this is the algorithm the site actually uses, but I'd bet it is.

Here are a few caveats to the above discussion.

Caveat 1. Everything here applies to the new-style SS lists (seen on the new (Beta) story reader). The old-style SS lists are not always identical to the new-style ones.

Caveat 2. The new-style SS lists are "series aware." Series are always referred to by their first chapter, and their ranking in the SS list seems to be based on the chapter with the largest number of fans in common

Caveat 3. The site does seem to do some filtering based on Categories. The one rule I'm pretty sure about is "I/T stories can't go on SS lists of non-I/T stories." This doesn't seem to be the case, though, for GM, LS, or Fetish. For example, I've seen a GM story on the SS list of a non-GM story. Whether or not there are other category filtering rules remains to be seen.

Caveat 4. The new-style SS lists are not updated immediately when a story is favorited. I have a story that was published in August, and its SS list seems to have only recently been updated from tag-based to favorites-based.

Caveat 5. If there are fewer than 20 favorites-based candidates, they will go into the top slots of the SS list, and the lower slots will be based on tags. As 8Letters and Bramblethorn have pointed out, sometimes there are not enough candidates of either type. In this case the SS list will be truncated.

Caveat 6. Some of the SS lists I've examined have slight discrepancies from the above rules that I can't currently explain.

Methodology. The number of fans in common between Story A and Story B can be determined by counting how many of the people who favorited Story A appear in the list of people who favorited Story B. The only way I know of to get the full list of people who favorited a story is to go to the last page of the story in the old (classic) story reader, have your browser show the page source code, and sift through the code for the list. I copied the lists into a spreadsheet to do the analysis for Story 5.

In order to verify that an SS list contains all the stories that have so many fans in common, you have to check every story on each fan's Favorites list. I've done this for some stories and generally found that their SS lists do contain all the stories most commonly favorited by their fans. This is not possible, however, if one of the fans is a "kazillion-favoriter." As Bramblethorn has pointed out, the Members Page truncates very long Favorites lists. As far as I can tell, the lists are truncated at 6000 entries, and they do not include the most recent ones. This makes it impossible to check the kazillion-favoriter's entire Favorites list.

Excellent work! This is what I suspected but it's nice to see real data as opposed to wild conjecture.
 
Excellent work! This is what I suspected but it's nice to see real data as opposed to wild conjecture.

Thanks. I should have mentioned that these results confirm Bramblethorn's and your previous conjectures.

Usually recommender systems are based on a "readers who liked X also liked Y" type of algorithm.

A quick and dirty method would look something like: select all the identifiable readers who loved the story (i.e. 5 stars and/or favourited), look at which other stories those same readers loved, and recommend the ones that show up most often. There are ways to refine it, but that's the basic idea.

I think the key determination of whether a story shows up on the Similar Stories list is looking at the number of readers that have favorited both stories. That seems most consistent to me with what I see on my stories' Similar Stories lists. It explains why the stories on these lists tend to have very high views, votes, and scores. It also explains why stories can show up on Similar Stories lists without being in the same category or sharing the same tags.
 
Back
Top