|By Avi Rosenthal||
|March 15, 2014 04:56 PM EDT||
If you think that you know better than others, you probably will not Adapt.
Example 1: A Bank
Few years ago, instead of receiving Web Page with one of my bank account reports, I received an error message.
As the report was consistently displayed correctly, it looks like a bug correction implemented with insufficient Regression Tests.
I notified the bank about the problem. They ignored my report and tried to find my incapability to use a Web Site (after more than 30 years in IT, including many years of using their Web Site).
After a week or two I identified the circumstances in which the report is not displayed correctly: It happens if and only if, the report should be displayed in more than one Web Page.
I send to the bank my analysis + the error message.
Unfortunately, I was not able to correct the bug without Source Code.
The result: Customers, including myself, had to wait two weeks for correction of a bug I could fix in an hour or two, if I could access the Source Code.
example 2: Another Bank
I found that a bank performed a transaction in a bank account I managed, without an order to do it. As the activity was an activity no customer would ask a bank to do, I thought that I had to go to talk to the clerk, who performed it.
Together we found the mistake the bank performed and corrected it. It was a mistake not an intention to steal money from the account.
However, based on my background in Computers Security Systems (Including Penetration Test in another bank), I easily found that someone (probably a bank employee) could use the same mechanism for criminal actions.
I even knew that an American bank programmer was convicted and jailed for stealing customers money using a similar mechanism to the mechanism used innocently in my account.
The Systems should be changed in order to identify unreasonable Orders and check them automatically or manually.
I wrote to the bank expecting that they will fix the vulnerability and thank me. I thought they probably pay me a little amount of money, which I intended to donate to an organization helping children suffering from cancer.
The bank thought that they new better: They ignored my message.
Do I post Spam?
I will list my writing experience and you will answer the question.
My writing activity includes:
1. An article based on my Masters Thesis in Psychology, written by me and my thesis moderator. He is a university Professor. The article was published in Hebrew in a local Scientific Journal.
2. Many articles on various Information Technology topics, which were published in local and European journals and local Web Sites.
3. Three blogs. The other two are written in Hebrew.
4. I participated in ebizq SOA Forum writing my opinion on questions asked.
5. I wrote about 100 articles in the Hebrew Wikipedia. A month ago I received a price for improving a stub article, during a competition for improving stub articles.
6. More than hundreds of posts in Facebook, Twitter, Googl+ and Linkedin (including many Linkedin groups).
Some of the posts received Comments, Likes and Shares.
Yesterday, I tried to post in Linkedin. I received a message that because someone (or some people) reported that a post of mine is Spam, the post will not displayed, but will be submitted to The Group Leader, who will check the post and decide to publish it or not to publish it.
What is the Problem?
There are two Linkedin Groups types:
1. Groups whose policy is to moderate posts prior to publishing.
The group leader is aware of posts he has to moderate. He will probably moderate the posts shortly after they were submitted.
As a member of many Groups of this type, I submitted posts which were published. Nothing was changed due to flagging a post of mine as Spam.
2. Groups whose policy is to publish automatically
The problem is that in those groups the group leader is unaware of Posts waiting moderation. According to Linkedin message it could take weeks until they approve my post and publish it.
Timing could be a significant factor for Posts Publishing.
Linkedin message send to me, recommended applying to Group Leaders and asking them to remove the restriction, so my posts could be published automatically.
I am a member of too many Groups of that type. I prefer to write this post instead of applying to all groups leaders.
It is not a personal problem. Any body's viable posts could be marked as Spam intentionally or unintentionally.
Is Linkedin Spam detection limited?
The answer to this question is Yes.
Linkedin should use the approach used by Anti-Spam Software, based on Bayesian algorithms.
If someone posted many posts and get Likes and Comments do not hurry to restrict him, just because one or few readers marked it as a Spam (intentionally or unintentionally).
On the other hand, if someone posted only 5 posts and many people marked every post as Spam stop him, by a mechanism similar to the mechanism used to stop me or even more extreme procedure.
It should be a gradual process based on additional information.
It looks like Linkedin's approach resembles the approach the Banks in example 1 and example 2 adopted.
According to the answer they send to me, when I suggested that they will improve their Spam detection algorithm, they stick to the current approach.
As anyone can mark any post as Spam, probably they may change their algorithm only after someone will intentionally mark as Spam many posts by leading contributors.
Nov. 30, 2015 12:00 AM EST Reads: 449
Nov. 29, 2015 06:00 PM EST Reads: 443
Nov. 29, 2015 02:00 PM EST Reads: 485
Nov. 29, 2015 01:00 PM EST Reads: 357
Nov. 29, 2015 12:45 PM EST Reads: 421
Nov. 29, 2015 12:30 PM EST Reads: 427
Nov. 29, 2015 12:00 PM EST Reads: 528
Nov. 29, 2015 11:45 AM EST Reads: 329
Nov. 29, 2015 09:45 AM EST Reads: 454
Nov. 29, 2015 09:15 AM EST Reads: 347
Nov. 29, 2015 08:45 AM EST Reads: 232
Nov. 29, 2015 08:00 AM EST Reads: 282
Nov. 29, 2015 07:00 AM EST Reads: 500
Nov. 29, 2015 06:45 AM EST Reads: 744
Nov. 29, 2015 06:00 AM EST Reads: 559
Nov. 29, 2015 06:00 AM EST Reads: 377
Nov. 29, 2015 05:00 AM EST Reads: 463
Nov. 29, 2015 04:30 AM EST Reads: 488
Nov. 29, 2015 04:00 AM EST Reads: 380
Nov. 29, 2015 03:00 AM EST Reads: 598