Last updated on April 1, 2019

Content Moderation vs. Spam: How to Control Repetitive User Posts

Controlling different spam contents

Spam contents are unwanted and unsolicited emails, instant messages or comments sent to people online in bulk. The purpose of spam is usually to advertise products or spread information. Otherwise referred to as junk mail or message, spam contents are usually encountered in the following:

  • Comment sections on public forums
  • Participating in a group chats
  • Filling out of survey or other forms that are not directed for emails.
  • Browsing/Interacting with a website without even subscribing to their mailings

 

Types of Spam

Most common types of spam contents

There are different types of spam, and some of these are most likely familiar to most internet users.

  1. Adult Content Spam – The aim of this form of spam is to advertise products to enhance sexual potency, linking to pornographic sites and/or advertising relevant links.

Example Ads are:

“Make your “P” get 10 inches larger using <Product name>”

“Do you want to satisfy your woman? Try <Product name> and make her feel the absolute happiness in bed”

This type of advertisements usually contains offers from dating agencies and porn sites.

  1. Health and Medicine – The purpose of health and medicine spam content is to advertise dietary supplements, skin care products, weight loss products, cure for baldness, posture improvements, and other traditional medications that are available online.

Example Ads are:

“Rejuvenate yourself and make your skin glow, use <Product name> and witness the magic”

“Do you want to make yourself 10 years younger? We have it here for you, use <Product name> and show the world the beauty of your skin”

  1. Computer Software and Internet – This category usually offers cheaper or discounted computer hardware and software and services for website owners. It also includes low price internet providers that are available online.

Example ads are:

 “Looking for cheap but quality <Software products>? We have it for you”

“Avail now! 20% discount for Internet, Call <Contact Number> and enjoy the fastest internet you’ll ever experience”

  1. Personal Finance – This category offers insurances, debt reduction services, and loans with low interest rates, to name a few.

Examples ads are:

“Make your family happy with establish finance – enjoy the lowest interest rates for loans here in <company name> and contact us at <contact number>

Generating leads and making money are the main objectives of spam. However, there are also several other reasons why people post or spread spam; and that includes spreading malware to unsuspecting users.
 
Particularly, there are spammers who try to convince users to reveal their personal information in order for them to steal and use it for illegal purposes.

 

Sneaky Methods Used By Spammers

Email icon with spam contents inside

Spammers will do anything to bypass online community guidelines and rules against spam. Although there are automated filters that monitor whether user posts are spammy or not, there are several methods and tricks used by spammers to outwit the policies and guidelines established online communities.

Below are some of the examples of sneaky methods spammers use:

  1. Using alphanumeric characters and symbols – By using a symbol equivalent to the word or message of the spam, automated filtering programs cannot detect the inappropriate content contained in the post.

Examples are: ( @ – # – $ – ) [email protected]$

  1. Pattern Recognition – Adding spaces in between letters or using alphanumeric characters to separate words to maintain spammy content without being detected by AI filtering.

Examples:  F R E E    F/R/E/E    F*R*E*E

  1. Context and Prediction– The human brain can determine and read words even when the letters are jumbled. For as long as the first and last letter remains, the human eye will still be able to recognize it.

Example:   VGAIRA SMPELS ONELIN

  1. Cloaking – By disguising the links added in spam ­– pornography, bogus products, and insurance scams which should be detected and flagged as inappropriate are regarded as safe and user-friendly. Cloaking spam becomes successful when Facebook’s moderation system overlooks the link and approves the content as one that is safe and adheres to its community’s policies.
  2. Blurring explicit images – AI moderators cannot detect the inappropriate content in an image. Disguised by the blur, explicit and spammy images will be made available for viewing across social media platforms and online forums.
  3. Adding Text to the image – A method that is a bit similar to blurring, adding text to images prevents automated moderation filters from detecting the inappropriate content in the manipulated photo.
  4. Creating fake accounts – To bypass the security system of social media, spammers create multiple accounts under fake names in order to continuously spread repetitive advertisements without being caught.
  5. Splitting words in HTML – One common trick spammers use is splitting words with HTML comments.

Example:

   Na<! – – 63 – ->ked   Gir<! – – adf  – – >ls

Would be rendered as:

  Naked Girls

  1. URL one liner – Spammers discovered that the best way to hide messages from AI filters is by not putting any text at all. Spammers send messages containing generic sentences containing a hyperlink. When the recipient clicks on the link, they will be taken to the landing page where the spammer introduces the product intended to be sold.
  2. ASCII “Art” – Another way to disguise spam is by using large letter glyphs to represent the message or the spam itself.

Example:

#########    ###########      ###########     ###########
#                       #                    #      #                            #
#                       #                    #      #                            #
#                       #                    #      #                           #
########      ###########     ###########     ###########
#                       #         #                #                            #
#                       #            #             #                            #
#                       #               #          ###########     ###########
 

 

Impact of Spam Contents on the Business Sector 

globe surrounded with spam contents

With the rise of spam contents, businesses around the world have been greatly affected, particularly in the areas of marketing, online branding and user experience. From the business owners’ point of view, leaving spam uncontrolled will lead their reputation and business to its major downfall.

Here are some of the most impactful negative effects of spam contents to businesses:

  1. Productivity/Profit Loss – If a business receives spam emails in bulk on a regular basis, it could add to the daily work load of the employees in charge of sifting through every single email sent to the company. This, in turn, wastes valuable work time. Wasted time equals significant loss in profit.
     
  2. Poses Potential Threat, and Legal Risk – If an employee is unaware of the sneaky methods used by spammers, as well as the risks that come with opening links found on spam, they could potentially jeopardize the company’s reputation.
     
    Even if the spam content found in a brand’s social media page or online community did not originate from the company or its employees, customer experience and satisfaction can diminish when users encounter spam.
     
    The worst case scenario is it may lead to the charging of fines, fees, and damage settlements especially if the spam content contains links to explicit websites and pop-ups of deliberate product selling.
     
  1. Backdoor of Malware Threats – Spammers uses Malware threats to steal sensitive and confidential information from the company and its end-users. The reason behind this is for spammers to drain their victim’s bank account or commit credit card fraud under their name.
     
    Businesses implementing the Bring Your Own Device Policy (BYOD) are among the most common targets of malware threats. Malware attachment received by an unsuspecting employee can cause major damage to the company once malware-infected files are received or downloaded.
     
    In other words, they could unknowingly spread virus across the entire company’s network simply from entertaining messages or advertisements considered as spam.

 

Controlling Spam through Content Moderation

spammers showing different spam content

To save the business’ reputation against spam contents, content moderation service is a must. Content moderation has become an integral part in the business world and the online community.
 
The role of content moderation is to control shared content from users and prevent users from opening unwanted emails, inappropriate, prohibited and repetitive user posts. It operates based on a set of guidelines, rules and restrictions that enumerate which content is allowed from what is not.

Listed below are some of the most suitable content moderation methods that can be used to control spam, and a couple of improvements that can be applied to boost user protection from spammers.

  1. Pre Moderation – Pre Assessment Moderation, despite not having the real-time capabilities of most types of moderation, is still the best option for businesses that are very particular about protecting their reputation. Below are the advantages that pre-moderation can offer:
    • Closely controlled community posts
    • Prevention of malware threats
    • Minimizes bullying, racism and discrimination
    • Faster prevention of unwanted sexual contents (especially for communities with children as its main audience)
    • Protection from legal risks
    • Maintenance of good reputation
       
  2. Reactive Moderation – The best thing about reactive moderation is it allows users to post freely without pre-assessment. Most social media users prefer this kind of moderation because moderators and end-users are actively participating in the common goal of eliminating inappropriate content posted online.
     
    Basically, the success of reactive moderation relies on the users reporting abusive content and the moderators monitoring and reviewing user posts that have been flagged or reported. Despite being commonly at risk of enabling the visibility of illegal contents, reactive moderation’s quality control against spam can still be enhanced through adding the following features:

    • Report Center – This section is where users are allowed to identify abusive contents and then notify moderators so that appropriate action can be implemented.
    • Profanity Filter – This section prevents divisive text that triggers heavily argumentative user posts and online bullying by filtering and/or censoring highly threatening comments and posts.
    • URL Filtering – With URL filtering, the user can blacklist suspicious-looking URLs or those that lead unsuspecting end-users to porn sites, pop-up ads, and even virus-infested files that could potentially harm their computer system.
    • Keyword Filter – A list of keywords can be specified and added in the moderation guidelines. These keywords contain the words, terms and/or phrases that are not allowed to be posted or made viewable to the brand’s online community. Once a spammer’s post triggers any of the given keywords, it can be easily detected by the moderators.
    • Warning System – This system will allow the moderator to monitor certain offenses on the user’s posts. The system may use warnings, suspension or total ban of the account depending on the degree of violation committed by the user.
       
  3. Automated moderation – It uses software or application to moderate abusive content. This type of moderation can be enhanced by improving the features that can detect and regulate spam. This includes:
    • Image Moderation tool – Image moderation automatically filters out abusive images posted online. This tool is usually accompanied and enhanced by features such as Optical Character Recognition application, Moderation User Interface, and Machine learning-based classifiers.
    • Text Moderation tool – Its role is to detect profanity or selectively disguised inappropriate words and languages on the internet. It used API (Auto Programming Interface), moderation’s text classification feature, and PII (Personally Identifiable Information) to detect the source of spammy contents or abusive posts.
    • Video Moderation tool – This particular tool detects adult and racy videos that are not appropriate for public viewing. This feature can be more enriching with the inclusion of Machine-assisted video classification, shot detection tool and human review tool to reject harmful video contents.
    • Human Judgment – Overall, every type of moderation software should always be paired with human judgment. Automated moderation may have its advantages and advancements, but it still has certain blind spots that only a human moderator can supplement. Combining the performance of an AI and a human moderator can uphold the safety of users against spammers more efficiently.
  1.  

Content Moderation as a tool for regulating spammy contents on the online community must be taken as a priority. Every business that wants to be safe from the harmful effect caused by spam should know that having the right and appropriate type of moderation is the best possible decision.

By combining the use of automated applications to moderate and human judgement to see and filter out the abusive contents stealthily used by spammers – success in moderating the online community will be achieved.

Get in touch with NMS

Get the Outsourcing Solutions you need through NMS! Fill-in the contact form below for any inquiries and we will get back to you as soon as possible.




Live Chat Offline

We are currently offline at the moment.
Please email us at
[email protected]
and we’ll get back to you in 24-48 hours.