I recently came across this script that extracts friend rank data from the currently logged-in Facebook profile and presents it as a table.
After trying the script personally, I became puzzled as to why certain individuals were consistently ranked higher than others. The rank seems to refresh daily, so I have experimented with various user interaction, and this shifts many entries appropriately; however, the same 'certain individual(s)' would often (with no discernible interaction) arbitrarily move up in rank.
My question is this: is it possible that this rank is being affected by other, external profile's usage data/habits?
In the interests of privacy, it seems very unlikely that anything but personal habits would influence this ranking, but my own and other peoples' usage anecdotes seem to suggest 'arbitrary' movement that would only be explained by external data.
I cannot seem to find a definitive answer to this elsewhere.
Any input would be greatly appreciated.
Related
Suppose player's scores determine their rank within a their division within a tournament. The, points are allocated according to their rank. The website hosts seasonal scores and tracks players cumulative points.
Thus, I am seeking a rewards system via php/mysql that would allow players to redeem their points for merchandise.
I have seen "points systems" for user interaction via web sites and only assume a similar system could be set up using player's performance points as described above.
I do not have PHP or DB coding abilities but feel this entirely possible. Has anyone seen a similar scenario as I have described? Just looking for thoughts and comments from those more knowledgeable than I on this subject. Thank you.
I, as well, felt it was possible. I have just have not seen any real world examples with similar circumstances. The "points and rewards" systems observed on the web seem to award points based on user interactions with the site rather than "on field" performance.
Possibly, the better question would be is where does one go to have such a system developed (allowing that I personally do not and will not have these abilities)??
I have built a recommender systems which has tens of thousands of items and their feature descriptions, but no user profiles as of now. I am looking for pointers to approaches that can help me bootstrap the system, so I can do some evaluation. I would appreciate any pointers to papers/applications that have addressed this problem.
How to deal with the cold-start problem depends a lot on your specific application.
An easy way of dealing with the user cold-start problem is to present the new user with random items, or the most popular items, or hand-selected items, and start learning from them.
Another way is to present users with a questionnaire, and then present items to them according to the results. Or you directly show them items/products and let them rate/select the ones they like.
Also note that in web-based system you usually know some things about your users: Which operating system/browser they use, where they (roughly) come from, which language they speak.
All this information can be used.
Papers:
see the Wikipedia article on the topic
My answer to another question on StackOverflow lists some papers for dealing with new items - most of the methods would also be applicable to new users.
Another approach is to select products/items that will help you most for learning about the user. Just out of my head, you can find them by querying Google Scholar for "recommendation" and the terms "decision trees", "active learning", "user cold-start", and so on.
When you're logged in, in the page source, there is a list called OrderedFriendsListInitialData.
According to rumour, it's a list of people that visit your profile the most, others say that it's a list of profiles you view the most, and yet other say it's the friends you interact with the most.
Can anyone shed some light on this by providing a definitive answer, or at least an educated one?
If You check the code You will notice it has something to do with right sidebar. Just before it in there is this url https://s-static.ak.facebook.com/rsrc.php/yT/r/q-Drar4Ade6.ogg
As it is JSON string obviously it has to be related, file this url is pointing at is sound notification for chat.
As You may notice it is initial data not chat list probably later chat script use this data to fill up people on list and make some extra check etc..
There is word ordered as well, Myself I'm not really active on facebook so have no way of checking it but it is known that fb analyses all Your steps and make this list based on thousands of factors to provide You with list of users You are likely to chat.
You father may be there because fb knows You are family and consider it as high possibility of conversation.
Send email to them If You want details.
Well in the time passed since you first posted this they've changed the name of the list to InitialChatFriendsList which I suppose is a little more descriptive of what it is, but as far as how they determine what to put on there I think my friends and I have come up with a very plausible explanation.
When determining who you are most likely to communicate with on their chat system, facebook will obviously use a whole number of factors weighted differently to determine who you most want to talk to and who you most need to talk to.
the most important is who you actually talk to... who on fb chat that you communicate with most frequently will obviously show up on your chat list.
who you have public interactions w/ (i.e tagging in at some location, picture tagging, actual wall comments etc.)
Now those two are two very large factors when determining who they put on your list, beyond that it is a combination of who looks at your page and whose page you look at. Based on my list and the list of my friends, we determined that if you are inclined to look at somebody else's page/posts a lot and they are likely to do the same for you, they will move up in rank even if you don't have an actual interactions on facebook. A couple of people who I would admit to "stalking" the most on fb are not even on my list (at least not on the top 50 which is where i stopped checking) while other people who I do occasionally look at and I have reason to believe they would be looking at my profile as well are fairly high on my list (around 10-15th place). And of course there are the completely random individuals who show up on the list who probably are stalkers.
Anyways, my point is there are so many factors that determine who is going to be on this list, you really can't just attribute it all to people who stalk you and people you stalk. While in for some people that would be the case, for most of the people on the list there is a whole list of reasons they're on there.
Of course this is all based on a very small pool of data, so who knows...
I think it may be the list of people who are on the top part of your chat list - the people you're statistically most likely to talk to. But! I may be wrong.
It definitely is the people who facebook considers are the most likely you are going to chat with. There are two lists of people in chat, one of the above, and the other friends who are online.
I believe the first 3 are accurate. When I checked for myself, my boyfriend was one, and my two best friends were 2 & 3. Everything after that seems to be a bit random, because #4 was a person I haven't interacted with for years.
I am pretty much sure that if you look carefully at any friend's timeline profile you can easily predict what going on in his/her life, Even you can write his/her entire life, you can also find out the hidden fact which he/she never told or updated directly but indirectly he/she shared n liked related thing which will help you to analyze his/her activity. Is it anyway possible to build an automated system which can read n analyze friends entire facebook profile, his/her shared stuff, likes, comments etc. and create a report which will expose his/her entire life facts including hidden one, using some AI or Machine learning concepts?
There's no system that will automatically be able to give content and understanding like you're looking for automatically. The human mind is able to infer a lot that computers simply can't understand. Also, you (generally) know some things about the people outside of Facebook (since you are friends with them) that fills in a lot of detail that the analysis system won't have.
The best thing you can do is to clearly define your problem and question that you're asking. There was a 'gaydar' project at MIT that was able to look at networks of students and generally correlate which ones are gay. For large groups you'll find it works overall, but for an individual person you're not going to be able to have great certainty.
Yet, to just ask the computer to 'find hidden information' won't work. You need to have a pretty solid model to work with. Overall, you're probably going to need a lot of data with confirmed facts to get started on testing that model as well (thousands of points needed). Also, with any social network you'll find that there is a lot of inaccurate/fake data on any given social network. People mis-list things all the time for various reasons (humor, etc) and this is going to throw off your models.
More specifically, what factors determine the priorities they assign in response to a given query? I'm looking for answers that address numerous scenarios including queries that...
Specify the "type" of result (objects such as users, posts, pages, etc. or connections like friendships, likes, tags, etc.),
Have authentication tokens as well as ones that don't.
Have conditionals such as "since" and "until."
Don't even specify a type, such as this search for the word query.
I am actually working on an app that uses /search to search places and I use a bit of all scenarios. I couldn't write down a specific order they appear in and to be honest I highly doubt it's something as easy.
I'm 99% sure it works like the Search in Facebook does, using the user data to bring up the most relevant results. I live in Ireland for 2 years now, but while testing the app I constantly receive search results from Romania and actually close to my Hometown, which are relevant to me.
Regarding your observations, Facebook's algorithms might take into account the source of the request as well - which would be good, means it only improves as your app gets more users.