Tutorial

[StrangerGithuber]

Hi
Very helpful the https://documentation.mosparo.io/docs/category/about documentation page however some of the part hard to understand. Maybe I am just stupid. I tried to find youtube video how to add rules and rulesets but no hope.
Is there video tutorial which explains and show how to do things? Extra point if there is test part too if the rules works.

Thanks in advance for the answers.

[zepich]

Hi @StrangerGithuber

Thank you very much for starting this discussion! You are not stupid! You want to use mosparo, which is a pretty wise decision... ;)

I'm very glad that you posted this question! For me, as a developer of this tool, it's sometimes hard to find the right amount of explanation in the documentation. Too much documentation, and nobody will probably read it, too little documentation, and people struggle. Because of this, I'm happy to read your post. Now I know we have to invest more time into this explanation, add more stuff like screenshots, and - yes - create HowTo videos to show how the functionality works.

Right now, we do not have video tutorials to explain the functions. We have that on our list of things we can do, but we haven't had the time to do it. I will create a HowTo video to explain how to create a rule in the next few days, and I would like to hear your feedback as soon as it is ready. For now, I would like to show you how to create a rule in this post.

How to create a rule

Introduction

Let's assume you received a submission that you do not want because it's spam or, in general, not interesting to you. This is the text of this submission:

Hello StrangerGithuber

Do you like watches? If yes, i have the best luxury watches in the world. Visit my store at www.best-luxury-watches.example.com.

Greetings

Luxury Watches Guy

To block a submission, we must determine which part of the submission is problematic. In this example, we have words like "luxury watches" and "watches" in general, and we have the URL "www.best-luxury-watches.example.com".

Steps to create a rule

1. Log into mosparo and choose your project from the project list
   
2. Now, you have to choose the rule type you want to use for this rule. The rule types have different use cases described here. We choose the rule type depending on the part of the submission we want to block. I'm choosing the Word rule type for this example to block the problematic words. Click on the rule type to create the rule.
   
3. Now, you see the interface to create the rule. In this interface, you have to enter a name for the rule. I'm using "Blocked words". If you want, you can add a description to explain the rule. You can also choose if the rule should be active or not, and you can enter a spam rating factor (more on that later).
   
4. On the right side, we can add one or multiple items to this rule. An item is, for example, one word, word combination, or regular expression pattern that should match. If you want to match multiple independent words, you have to add multiple items (one per word). We want to block the words and word combinations 'Luxury watches', 'luxury', and 'watches', so we add these as items.
   
5. In the field Rating, we can specify a rating for each item. By default, all items have a rating of 1, which means that the submission gets 1 point if one of these words is used in a submission. Since we want to block these words, we can increase the rating and set a different number. In my example, I set the rating for the luxury watches to 5 because that's always spam for me. But the words luxury and watches are not directly spam, so I set them to 2.
   
6. Now, we can save the rule.
   

We've officially created a new rule. Now, the next submission with these words is blocked.

Steps to test if the rule works

1. Go to Tools -> Rule tester
   
2. Enter the text of the submission (see the box above with the example text) in the Test value field and choose Text field (Multi-line) in the dropdown below (Data type).
   
3. You can choose to test only rules or rulesets or both together. It's normally recommended that you test both together.
4. Click the button Test in the bottom right corner
   
5. mosparo now tested the given content with the available (and selected) rules and rulesets. First, you see the content you've entered. Below, you see the Detection result. In this example, the created rule would catch this submission if the user submitted the same content again. You see that mosparo identifies this submission as spam with a spam rating of 9. The 9 is calculated because the rule items luxury watches with 5 points, the item luxury with 2 points, and the item watches with 2 points were found, so 5 + 2 + 2 = 9.
   
6. The spam score on the right side is the number of points a submission needs to be qualified as spam. You can adjust that value in the project settings if needed.
   
7. Below these values, you will see the matched rules and the rating of these rule items. A rule item that is not matching would not be listed in this list.
   

What is the Spam rating factor?

With the spam rating factor (the last field on the left side when creating a rule), you can increase the rating of all rule items at once. Let's imagine we set the spam rating factor of our test rule to 3, and then it will multiply the ratings of our rule items with this factor.

For the rule item luxury watches, the test will result in 15 points (5 * 3), while the other two rule items will have 6 points (2 * 3). With these spam rating factors, it's easy to increase or decrease (if you use a number below 1.0, for example, 0.5) the ratings for all items in a rule. With a spam rating factor of 3, you will have this result when testing the rule:

Rulesets

The idea of rulesets is to use a shared list of rules from the internet. To be able to use such a ruleset, someone has to create a ruleset first. As far as I know, there are no publicly available rulesets right now, but maybe later. To add a ruleset to mosparo, you need the URL for this ruleset before you can add it.

The format of a ruleset is described here. I don't know why, but it is accidentally described as 'Rule packages' instead of 'Rulesets'. But a ruleset, in general, is a JSON file with the format as described in the documentation.

You can use the URL https://zepi.io/dev/mosparo/ruleset.json for a test, but that's a simple test ruleset that is not really made for spam detection.

---

Does this explanation better help you understand how to create a rule? Or do you need more details or explanation? If this helps you, I would use it to create a HowTo video.

Thank you very much for your help in making mosparo better!

Kind regards,

zepich

[StrangerGithuber]

Hi @zepich ,
Thanks for the kind words and holy moly macaroni thanks for the detailed information with images. Indeed easy to understand this way.
I mentioned videos because:

1. Easy to understand especially when the person show the test result too
2. Good advertisement material so more people will be reached.

Personally when I searched for "laravel spam protection", "spam protection locally", "laravel captcha", "local captcha" and so on then in the 3rd page came up this project (laravel because I am using laravel framework to create websites). When I checked on youtube then just illustrations exists how this project works but no real action how to use. Ruleset is not that popular maybe because not many people knows about this ( I don't have statistic how many people knows about this project but from youtube and hard to find rulesets shows that not enough people did reach this project ).

I am planning to make videos too when I will understand fully this project because so far looks promising at least I could help the project like this and today I will help with translation.

Just now I have hard time to connect my laravel frontend with this project because everytime I get error because I get this error:

Access to XMLHttpRequest at 'https://captcha.network/api/v1/frontend/request-submit-token' from origin 'https://help.network' has been blocked by CORS policy: No 'Access-Control-Allow-Origin' header is present on the requested resource.

Laravel web is help.network and this project is captcha.network. Both are in localhost however using local dns so just I can reach.
I added to the "Hosts" but it doesn't work. Maybe something else I need to do too, not just that.

Update: Ok I found the issue with this problem. Server settings error.

[zepich]

Hi @StrangerGithuber

Thank you very much for your message! I'm happy to hear that!

You're absolutely right. Videos are a good option for being more visible. We have to invest more time into them.

Regarding the CORS error message:

1. You configured the hosts correctly. For a test, you could also define a host with simply the wildcard * without anything else to check if that works.
2. We had cases in the past where people had multiple projects and accidentally connected to the wrong project (wrong UUID in the frontend settings). Can you confirm that that's correct?
3. If the host is configured correctly (as in your case) and the correct project is used, the third problem could be technical issues. Can you have a look at the file var/log/prod-2025-01-16.log? Can you post the content or at least the last lines of this file? If it does not exist, this is not the problem for the CORS header.
4. It's also important that all sites use SSL, but as far as I can tell, that's already the case in your setup.
5. Are these two things on the same web server, or are these different web servers?

Thank you very much for your feedback.

Kind regards,

zepich

[StrangerGithuber]

Yeah I found the issue. It was on my side. Server settings were not right.
Thanks again for your hard work. :)

[zepich]

Perfect, thank you very much for your feedback!