Inspiration

With billions of users online, filtering out offensive content ranging from pornography to promoting graphic violence and extremism is a never-ending task.

Moderating inappropriate content online is really essential to protect innocent users from experiencing horrific content. And, for the most part, this job largely falls on teams of staffers who spend most of their days sifting through offensive content manually.

But, what if we can have an automation service that can be used it to augment human moderation of environments.

Environments where partners, employees and consumers generate text content. These include chat rooms, discussion boards, chatbots, eCommerce catalogs, and documents.

One in which it can flag text that may be deemed inappropriate depending on context.

What it does

Now this is possible with Text Moderator Activity for UiPath. This will help you detect potential profanity in more than 100 languages and match text against your custom lists automatically. Text Moderator activity also checks for possible personally identifiable information (PII).

enter image description here

enter image description here

  • Detect potential profanity in text with term-based filtering

If the API detects any profane terms in any of the supported languages, those terms are included in the response. The response also contains their location (Index) in the original text. The ListId in the following sample JSON refers to terms found in custom term lists if available.

Content Moderator’s machine-assisted text classification feature supports English only, and helps detect potentially undesired content. The flagged content may be assessed as inappropriate depending on context. It conveys the likelihood of each category and may recommend a human review. The feature uses a trained model to identify possible abusive, derogatory or discriminatory language. This includes slang, abbreviated words, offensive, and intentionally misspelled words for review.

  • Use machine-learning-based models to classify the text into three categories.
  • Detect personally identifiable information (PII) such as US and UK phone numbers, email addresses, and US mailing addresses.

The PII feature detects the potential presence of this information:

Email address US Mailing address IP address US Phone number UK Phone number Social Security Number (SSN)

  • Normalize text and autocorrect typos

  • If you ask for auto-correction, the response contains the corrected version of the text:

How I built it

The project was built using the following technologies.

  • Visual Studio with .NET Framework 4.6.1
  • Azure Cognitive Services - Advanced intelligence APIs harnessing the power of Machine Learning

Challenges I ran into

I encountered issues finding a away how I could best present the output evaluation data in a way that it is 100% complete (both for developers and functional users) but also considering not to overwhelm end users with unnecessary information.

Accomplishments that I'm proud of

When it comes to me that I was able to create even greater value out of products that’s already providing the best value! I'm really proud to have developed another channel for this wonderful solution/API to be utilized. One in which the automation community can get the hands of and build really intelligent automation workflows that can accelerate businesses' journey to digital transformation.

What I learned

But seriously, not only I learned a lot of technical stuff interacting with cognitive services built within Azure cloud platform, but also I learned a lot about how I could really create more value out of a product that already is providing a lot of value from the start.

What's next for Intelligent Activities - Document and Text Translation

More intelligent activities to come!

Built With

Share this project:
×

Updates