AI Zone Admin Forum Add your forum

NEWS: Chatbots.org survey on 3000 US and UK consumers shows it is time for chatbot integration in customer service!read more..

Smart forum Scoring Proposal, to Reward Quality not only Quantity of the member’s work
 
 

I just want to suggest something for the forum.

The “membership score” should be different, now it measures activity, has yet no grounding on how fast the member react, on what quality of text he post, even on the agreement of all the forum’s community.

As I like to make hybrid measures, who may represent better the “score” or a forum-poster (many social measuring theories are about this) I make a suggestion:

Add some grades 1..5 “recommend” or “I like this post”  button including “I don’t like this post” (check http://www.CodeProject.com, it has a good measurement system)

Even you can add some word, to describe the post-quality to letter include the more frequent as a drop-down box to tag the posts better.

The amount of influence on the “post score” and his author’s score should be proportional to the “voter’s” score also, it’s no use that a newbie like me do a harsh critic on a good post from another, nor that a ‘smart-ass’ who always do silly critics and swears around (raising actually his score) knocks down a good reflection or thought.

This is simple math, I can help with the modelling (contribute to it), despite don’t eliminate the actual score, just change the “outcome” like senior/etc. the internal score may not be show able, but the influence and the quality arising on this interaction will be astonishing good, believe me!

This will lead you to classify better the posts, letting the newbie (like me) navigate accurately into the more interesting (voted) threads.

Also the classification not is like a tree (as any other forum), and it collapses fast in meaning-coherence if it is not well designed (like I think this is) don’t panic!,

But it doesn’t lack of design, its a pitfall of the human mind!

So I can make (at least try) to make some suggestion to help in classifying the posts automatically, aided by a social motor.

Here is the core idea:

You start not from scratch, but from the actual labels, add them a “explanation” (moderated but with suggestion-ability) and put them into a drop-down box, with clicks on each item (select-ability), or arrange them dynamically on top of each post, with a clickbox. Then each member who reads the post, may vote and even may vote for a change (or addition) of the classifications of this post, which may not be unique, like humans think (multi-classified). Leaving also a box to add a suggested category, and explanation of it (to avoid ambiguities, like a synset-sense in Wordnet) moderated of course, So ¿What are the benefits?

You can then navigate in many dimensions, inside the post, and to “similar” categorical posts, chronologically ordered, and even collapsing the “least relevant” posts.


Also you may navigate from the top of the woods, not by tree branch but in an ‘intelligent’ way, by letting the user type the subject(s), and do a semantic-(relevance+pertinence) weighted search inside the threads, and present him the best posts on this subject, to jump right into the best+most-voted post.

Even better, you may filter each post using NLP (very little) and make a information extraction, like keywords, so you may also classify automatically the posts based on the user-classification, but this is more complicated but I think its worth, so may be a user may navigate a cloud tag, instead of a post-thread.

I share this ideas with you to contribute a little bit, because I did it inside my bot’s logical mind-structure, to dynamically classify the dialog-actions of prior prior conversations, and the result are astonishing! hope here this can yield the same outcome!

Hope this helps wink best regards.

Andrés Hohendahl

 

 
  [ # 1 ]

Hi Andrés,

I LOVE your ideas and I really appreciate all the time and effort you took to note this down so precisely!

Without going in every single detail at the moment (I’ll do that when we’re about to implement version 2.9 which contains serious forum upgrades), I’d like to summarize your ideas to the essential, to make sure I understand what you’re saying.

In the core, you’re trying to create a ‘relevance engine’: what is relevant to the member? It’s based on additional action by members, like tagging & voting, and revealing it by smart navigation systems like cloud tags.

In this approach we need to reward people to spend extra time on tagging and voting. History has shown that most of the members are not prepared to put a lot of work in it, that’s why the Facebook Like button is so popular. It is just pressing ‘Like’ and that does the job.

In order to cover this I’d like to add the social elements. In the account section, we’ll ask members to connect their Chatbots.org accounts to their Facebook accounts. This will allows us to see who their friends are, and more importantly, which of those friends are registered on Chatbots.org. We’ll add a Like button + a tag field to each forum and content entry (thus all other tabs as well), probably with a ‘Chatbots.org bar’ at the very bottom of the browser window. This bar also shows: ‘(your friend) Andy just tagged page ‘xxx’ with ‘tools’ or ‘inspiring’ or ‘model’).

The bar will also show the general tag cloud, but also the tag cloud constructed by your (friends of) friends (or colleagues by connecting to LinkedIn, our research peers by connected through Academia.edu).

In the idea situation, we could even use NLP to dig even more deeper. That something we’d like to cover in version 3.0 of the website, where we’ll introduce a feature called ‘supersearch’, which searches through all areas. We must also introduce the Chatbots.org chatbot during these days. This will take a while though…..

What would you think of this addition?

 

 
  [ # 2 ]
Erwin Van Lun - May 3, 2011:

Hi Andrés,

I LOVE your ideas and I really appreciate all the time and effort you took to note this down so precisely!

Thank you, I’m flattered!

Erwin Van Lun - May 3, 2011:

Hi Andrés,
What would you think of this addition?

BEWARE
I must say, sincerely: I don’t go with the crowd as well as all new ‘things or concepts’ must be first tested exhaustively, and time and user number are a critical fact. Facebook & Tweeter services are gossip-like services, their fast success may be more on some human flaws than on the positive side of friendship. LinkedIn is a better service, indeed.

THE PROBLEM I SEE
What I criticize the most is that the social-services are the unique links: “aka: friend-of..” which are not categorized, and you may not be ‘friend of’ a customer you barely heard of but asked your friendship, you also are not friend of a company… (nonsense)
But for Facebook all ‘accepted-links’ are friends! or followers! and this is holy crap! (no one can have 100+ friends in his whole live and know each other) you may know or have accidentally met some of those people, but the contact may happened only once in a long time or even never!. Also there are several dangerous and inexact transitive ‘inferences’ made by the social hubs that are false or halve-truths: a friend of a friend is not always my friend! Indeed, they make too many dangerous assumptions, and many ‘commercial-enemies’ make smart-use and intelligence by sending someone else to be your friend and ‘steal’ all of your relations, to sell them something.. or even worse, to blame you on some thing, making severe misuse of the relations..
The misuse capabilities of such an openness in social hubs is still on it’s childhood, and as you will see, the community will still claim for more strict ‘information security’.
This happened on Facebook last year where they got too much ‘attorney’s’ calls and they was sued more than once for this kind of things. Even Google an Yahoo got sued and enforced to pay a fee for providing a ‘bad link’ to porno sites and photography sites, when you searched for some ‘movie star’.

BUT
This nevertheless collides with that you ‘may have’ a public profile (as mine, academically) but you really should know ‘who’ may get access to your sensitive data, and when you have 200+ contacts this fact becomes an utopia, you will (for sure) loose the control absolutely and you never will tell enemies from friends any more. And to defend on this, you probably would make more different ‘public profiles’ but again!, to maintain 2 or more is overwhelmingly complicated, so you are stuck!. Many busy people, after this experience may simply abandon the social hubs.

BEWARE
Also when you give access to a social-enabled application, like Facebook, they (the company behind the app) will get to know ‘who’ are your contacts, (its like you give open access to all your phone-book) and you may not want this to happen (like myself) because you don’t know what implication this may have in a near future. Many people don’t think on this, and are lazy, they may suffer on consequences sooner or later, but other simply won’t access those services (I never do) any more, obscuring their sole creation.

MY CONCLUSION
I think the social-network should only carefully used, be conservative in how to use it, explore the ‘good things’ of the most active members, reward the good ones and get rid of the evil ones. Make a community grow is show them what is good and what is wrong, its better to have an active community of 500 members than a roman-Coliseum with 50.000 killing each other on the arena. Even if the ‘business’ grow as resulting on advertising and banners. Its fairly good that many people use a site, but it’s far better that there may are less people making a quality use of the site, and get charged for this.
Use only ‘anonymous’ statistical and tendency information, like (43% of your friends recommend this guy/site/service/product) but not “John and Mary likes it’ because it may create dangerous implications, you will never know! but the future will surely tell!

Publicity and massive approaches, blended with Ads are tempting, they seem to be free, but as you grow old and get the experience of lifetime: things will get clearer and you will get to know that “there is no free lunch” and that time is your only asset on life and spending it ‘mocking on people’ or ‘reading gossip’ on Tweeter or Facebook is not a useful nor smart way to spend this precious and unique ‘coins’ the live gave you, aka: time

MANY MISUSES
Therefore I suggested on my last posts, to make ‘smart’ scoring on the posts, and the posters, to wast less of the time of ‘good intended’ people in reading bullshit!. Most of the forums I know are plagued with gossip-mocking-and blaming stuff, and this is worse when the moderators don’t react on this promptly and effectively. Your forum has a rather good balance, but for my delights, it lacks a good approach on finding things inside. The search engine don’t work at all: (as I posted some things on many threads, I wanted to find a word I have posted ‘‘million’, and it never found it!).. revise it for the next generation.

Some of the best forums are built on people having common interests, engineers, programmers (like here) among other technical guys.

I GOT AN IDEA
As this is a chat-bot forum/site. ¿why not put a chatbot to answer things about the site, and the forums content? Helping each visitor in seeing the most accurate posts/information/etc.
Even this chatbot, may contact each of the ‘users’ by way of a MSN/ICQ/IM link and ‘post’ them the news, online, In a similar way like the ‘alert’ mails…Even they may chat to each other, on a second ‘linked’ method.

I think that I can help on this with my ‘smart’ engine..

best regards
Andres

 

 
  [ # 3 ]
Erwin Van Lun - May 3, 2011:

Hi Andrés,

I LOVE your ideas and I really appreciate all the time and effort you took to note this down so precisely!

Thank you, I’m flattered!

Erwin Van Lun - May 3, 2011:

Hi Andrés,
What would you think of this addition?

BEWARE
I must say, sincerely: I don’t go with the crowd as well as all new ‘things or concepts’ must be first tested exhaustively, and time and user number are a critical fact. Facebook & Tweeter services are gossip-like services, their fast success may be more on some human flaws than on the positive side of friendship. LinkedIn is a better service, indeed.

THE PROBLEM I SEE
What I criticize the most is that the social-services are the unique links: “aka: friend-of..” which are not categorized, and you may not be ‘friend of’ a customer you barely heard of but asked your friendship, you also are not friend of a company… (nonsense)
But for Facebook all ‘accepted-links’ are friends! or followers! and this is holy crap! (no one can have 100+ friends in his whole live and know each other) you may know or have accidentally met some of those people, but the contact may happened only once in a long time or even never!. Also there are several dangerous and inexact transitive ‘inferences’ made by the social hubs that are false or halve-truths: a friend of a friend is not always my friend! Indeed, they make too many dangerous assumptions, and many ‘commercial-enemies’ make smart-use and intelligence by sending someone else to be your friend and ‘steal’ all of your relations, to sell them something.. or even worse, to blame you on some thing, making severe misuse of the relations..
The misuse capabilities of such an openness in social hubs is still on it’s childhood, and as you will see, the community will still claim for more strict ‘information security’.
This happened on Facebook last year where they got too much ‘attorney’s’ calls and they was sued more than once for this kind of things. Even Google an Yahoo got sued and enforced to pay a fee for providing a ‘bad link’ to porno sites and photography sites, when you searched for some ‘movie star’.

BUT
This nevertheless collides with that you ‘may have’ a public profile (as mine, academically) but you really should know ‘who’ may get access to your sensitive data, and when you have 200+ contacts this fact becomes an utopia, you will (for sure) loose the control absolutely and you never will tell enemies from friends any more. And to defend on this, you probably would make more different ‘public profiles’ but again!, to maintain 2 or more is overwhelmingly complicated, so you are stuck!. Many busy people, after this experience may simply abandon the social hubs.

BEWARE
Also when you give access to a social-enabled application, like Facebook, they (the company behind the app) will get to know ‘who’ are your contacts, (its like you give open access to all your phone-book) and you may not want this to happen (like myself) because you don’t know what implication this may have in a near future. Many people don’t think on this, and are lazy, they may suffer on consequences sooner or later, but other simply won’t access those services (I never do) any more, obscuring their sole creation.

MY CONCLUSION
I think the social-network should only carefully used, be conservative in how to use it, explore the ‘good things’ of the most active members, reward the good ones and get rid of the evil ones. Make a community grow is show them what is good and what is wrong, its better to have an active community of 500 members than a roman-Coliseum with 50.000 killing each other on the arena. Even if the ‘business’ grow as resulting on advertising and banners. Its fairly good that many people use a site, but it’s far better that there may are less people making a quality use of the site, and get charged for this.
Use only ‘anonymous’ statistical and tendency information, like (43% of your friends recommend this guy/site/service/product) but not “John and Mary likes it’ because it may create dangerous implications, you will never know! but the future will surely tell!

Publicity and massive approaches, blended with Ads are tempting, they seem to be free, but as you grow old and get the experience of lifetime: things will get clearer and you will get to know that “there is no free lunch” and that time is your only asset on life and spending it ‘mocking on people’ or ‘reading gossip’ on Tweeter or Facebook is not a useful nor smart way to spend this precious and unique ‘coins’ the life gave you, aka: time

MANY MISUSES
Therefore I suggested on my last posts, to make ‘smart’ scoring on the posts, and the posters, to wast less of the time of ‘good intended’ people in reading bullshit!. Most of the forums I know are plagued with gossip-mocking-and blaming stuff, and this is worse when the moderators don’t react on this promptly and effectively. Your forum has a rather good balance, but for my delights, it lacks a good approach on finding things inside. The search engine don’t work at all: (as I posted some things on many threads, I wanted to find a word I have posted ‘‘million’, and it never found it!).. revise it for the next generation.

Some of the best forums are built on people having common interests, engineers, programmers (like here) among other technical guys.

I GOT AN IDEA
As this is a chat-bot forum/site. ¿why not put a chatbot to answer things about the site, and the forums content? Helping each visitor in seeing the most accurate posts/information/etc.
Even this chatbot, may contact each of the ‘users’ by way of a MSN/ICQ/IM link and ‘post’ them the news, online, In a similar way like the ‘alert’ mails…Even they may chat to each other, on a second ‘linked’ method.

I think that I can help on this with my ‘smart’ engine..

best regards
Andres

 

 
  [ # 4 ]

Social: we’ll take all your ideas into account and we’ll listen to all members and their requirements. It will never be a mandatory feature, you can simply use Chatbots.org without connecting to any of the social networks.

AndyHo - May 3, 2011:

I GOT AN IDEA
As this is a chat-bot forum/site. ¿why not put a chatbot to answer things about the site, and the forums content? Helping each visitor in seeing the most accurate posts/information/etc.
Even this chatbot, may contact each of the ‘users’ by way of a MSN/ICQ/IM link and ‘post’ them the news, online, In a similar way like the ‘alert’ mails…Even they may chat to each other, on a second ‘linked’ method.

The holy grail of Chatbots.org grin. We’ll introduce a ‘stupid’ avatar shortly (though nicely integrated), and work on a intelligent variant next year. \
However, we’ll need to focus on creating revenues first, as we’re working on this project full time, and hardly any income, and we have to pay our houses, foods and cars as well). Do you happen to know professional developers in Argentina (or in wider Latin America) who would be willing to participate in our advanced members program and are not listed on our companies page:
http://www.chatbots.org/companies

All suggestions would be very appreciated!

 

 

 
  login or register to react