- Compatible XF 1.x versions
- 1.4, 1.5
- 1 site per license
- Updates duration
- 12 Months (£10 Yearly Renewal)
- Visible branding
- Yes, but removable with payment
What it does
Conversation Monitor scans new private conversation messages for specific keywords. If it finds a match, it puts the message in the moderation queue and creates a thread in a forum showing details of the conversation message.
Why it’s useful
For AVForums, it gives an extra layer of protection from classifieds scammers. But it could also be used for a number of reasons where people return to your forums to post similar messages to people and you need to stop them before the messages arrive. To alert admin if someone is spamming a particular website, for example.
An example of its use
Having operated classified adverts on our forums for years, we have found that one of the distinct advantages we have over ebay etc. is the sense of community of a forum which makes people feel safe to trade.
It is paramount that we do everything we can to maintain this safe environment.
Unfortunately, the more popular our forum has become, the more we are targeted by people who want to make a fast buck at the expense of our trusting community.
Of course, we permanently ban scammers, but they are often persistent and get round our multiple login detectors to re-register and scam people again.
Since the scammers share their private details with their intended targets by private conversation (such as bank account numbers, sort codes, post (zip) codes and phone numbers) we have for a long time wanted a way to monitor private messages and be alerted when a specific key word such as a bank account number is sent.
We have therefore commissioned this addon to monitor conversations for specific key words.
In just a couple days running on AVForums, it flagged up two scammers sending bank account numbers to other members to make payments. It enabled an admin person to alert the recipients and advise them not to send any money. We stopped two members (at least) from being scammed, found two known scammers with new accounts and banned them.
Set up – Admin options
The admin options allow you to choose which forum you want the threads to be created in and the user name of the thread starter.
There is also a usergroup permission to bypass monitoring so that your moderators and admin can discuss scammers without being flagged.
Setup – Keywords
Applications -> Conversation Monitor -> Keyword List
Each ‘keyword’ is actually a set of key words and associated known users.
So for example, we might have several accounts which have been used by a known person which have used a few different bank accounts and/or mobile phone numbers and/or post (zip) codes.
We set up a Title for our keyword set, such as the real name of the scammer involved or their most well known account name amongst the moderators.
We then enter up to ten different key words, each with an option to match exactly or not.
An exact match requires a literal match which you might use to monitor for an address with spaces.
A non exact match will strip away all characters that aren't a-z, 0-9, @, ., (, or ) to, for example, monitor for a phone number with or without spaces.
Message: This is 9 4 1.
Exact = No Match. Non-Exact = Match.
We also have a description field which allows us to enter, say, the last known address of the scammer or a thread URL where they are discussed.
We have a list of associated users names used by the scammer.
And lastly we have an option to enable this keyword set.
The Alert Thread
The thread produced when a keyword is found in a new message shows
- The sender
- The recipient
- The conversation
- The matched keyword(s)
- The associated users
Applications -> Conversation Monitor -> Log
Lastly there is a log of the matched keywords.
All the monitoring is done in a deferred task after the conversation is submitted. The user will not notice any delay.