Skip to content

OTS 1.7

Choose a tag to compare

View all tags
@ajamous ajamous released this 18 Mar 13:33
· 629 commits to main since this release
1_7
76396ee

Release Notes for OTS Version 1.7

We are thrilled to announce the release of OTS Version 1.7, which brings several significant updates to our platform. This version focuses on expanding language support and enhancing multi-language processing capabilities. Here are the key features and improvements included in this update:

Highlights

  • mBERT Integration for Multi-language Support: In our commitment to supporting a wide range of languages, OTS 1.7 introduces the integration of mBERT (multilingual BERT). This powerful model enhances our platform's ability to understand, process, and generate content across multiple languages more effectively. With mBERT, OTS now offers better support for language-specific nuances and a broader understanding of global languages.

  • Full Arabic Dataset Integration: We have fully integrated a comprehensive Arabic dataset to expand our coverage and improve processing, understanding, and insights for Arabic text. This enhancement aims to cater to our diverse global audience, offering better support for Arabic language content.

  • Indonesian Spam/Ham Dataset Addition: A new Indonesian spam/ham dataset has been added to improve our spam detection capabilities in Indonesian text. While this initial dataset is small, it marks a strategic enhancement to refine spam detection accuracy. Users can expect an expanded dataset in version 1.8.

Coming Soon

  • Expanded Indonesian Dataset: Our commitment to enhancing dataset accuracy and comprehensiveness continues. In version 1.8, we aim to significantly expand the Indonesian dataset with more extensive spam/ham/phishing data, further improving our spam detection capabilities.

  • Add Russian Dataset: Our commitment to enhancing dataset accuracy and comprehensiveness continues. In version 1.9, we aim to add the Russian dataset with more extensive spam/ham/phishing data, further improving our spam detection capabilities.

Examples

In the latest release notes for OTS Version 1.7, we're excited to share examples demonstrating the power and efficiency of our newly integrated features. Below, you'll find cleaned-up examples showcasing how to interact with OTS using the command line for spam detection in multiple languages, particularly highlighting our advancements in Arabic language processing with the mBERT model.

Example 1: Detecting Spam in Arabic

This example demonstrates how to identify a typical spam message in Arabic, promising a large cash prize.

Request:

curl -X 'POST' \
  'http://localhost:8002/predict/' \
  -H 'accept: application/json' \
  -H 'Content-Type: application/json' \
  -d '{
  "text": "مبروك! لقد فزت بجائزة قدرها 10,000 دولار. للمطالبة بجائزتك، يرجى زيارة موقعنا الإلكتروني الآن.",
  "model": "bert",
  "bert_version": "bert-base-multilingual-cased"
}'

Response:

{
  "label": "spam",
  "probability": 0.7167165875434875,
  "processing_time": 0.10480880737304688,
  "Model_Name": "OTS_bert",
  "Model_Version": "bert-base-multilingual-cased",
  "Model_Author": "TelecomsXChange (TCXC)",
  "Last_Training": "2024-03-11"
}

Example 2: Recognizing Ham (Legitimate Message) in Arabic

This example shows the platform distinguishing a legitimate customer support message in Arabic.

Request:

curl -X 'POST' \
  'http://localhost:8002/predict/' \
  -H 'accept: application/json' \
  -H 'Content-Type: application/json' \
  -d '{
  "text": "مرحبا، كيف يمكنني مساعدتك اليوم؟ نحن هنا لدعمك في استفساراتك.",
  "model": "bert",
  "bert_version": "bert-base-multilingual-cased"
}'

Response:

{
  "label": "ham",
  "probability": 0.9993540644645691,
  "processing_time": 0.11941719055175781,
  "Model_Name": "OTS_bert",
  "Model_Version": "bert-base-multilingual-cased",
  "Model_Author": "TelecomsXChange (TCXC)",
  "Last_Training": "2024-03-11"
}

Example 3: Another Spam Detection Example in Arabic

Highlighting another spam message that entices users with a special prize through a suspicious link.

Request:

curl -X 'POST' \
  'http://localhost:8002/predict/' \
  -H 'accept: application/json' \
  -H 'Content-Type: application/json' \
  -d '{
  "text": "تهانينا! لقد تم اختيارك للفوز بجائزة خاصة منا. اضغط هنا للمطالبة بجائزتك الآن: http://tinyurl.com/exampleurl",
  "model": "bert",
  "bert_version": "bert-base-multilingual-cased"
}'

Response:

{
  "label": "spam",
  "probability": 0.9766722917556763,
  "processing_time": 0.1628100872039795,
  "Model_Name": "OTS_bert",
  "Model_Version": "bert-base-multilingual-cased",
  "Model_Author": "TelecomsXChange (TCXC)",
  "Last_Training": "2024-03-11"
}

These examples are just a glimpse of the capabilities of OTS 1.7, especially with our efforts to enhance support for global languages through the integration of mBERT and the addition of comprehensive datasets.

What's Changed

  • Bump fastapi from 0.103.2 to 0.109.1 in /src/BERT/training/ots-apple-silicon by @dependabot in #9
  • Bump cryptography from 39.0.0 to 42.0.4 in /src/BERT/training/ots-apple-silicon by @dependabot in #8
  • Bump fonttools from 4.38.0 to 4.43.0 in /src/BERT/training/ots-apple-silicon by @dependabot in #7
  • Bump pillow from 9.4.0 to 10.2.0 in /src/BERT/training/ots-apple-silicon by @dependabot in #6
  • Bump aiohttp from 3.8.3 to 3.9.2 in /src/BERT/training/ots-apple-silicon by @dependabot in #5
  • Bump starlette from 0.27.0 to 0.36.2 in /src/BERT/training/ots-apple-silicon by @dependabot in #4
  • Bump werkzeug from 2.2.2 to 2.3.8 in /src/BERT/training/ots-apple-silicon by @dependabot in #10
  • Bump urllib3 from 1.26.14 to 1.26.18 in /src/BERT/training/ots-apple-silicon by @dependabot in #11
  • Bump jinja2 from 3.1.2 to 3.1.3 in /src/BERT/training/ots-apple-silicon by @dependabot in #3
  • Bump markdown-it-py from 2.1.0 to 2.2.0 in /src/BERT/training/ots-apple-silicon by @dependabot in #14
  • Bump requests from 2.28.2 to 2.31.0 in /src/BERT/training/ots-apple-silicon by @dependabot in #15
  • Bump flask from 2.2.2 to 2.2.5 in /src/BERT/training/ots-apple-silicon by @dependabot in #13
  • Bump pygments from 2.14.0 to 2.15.0 in /src/BERT/training/ots-apple-silicon by @dependabot in #12
  • Bump scipy from 1.9.3 to 1.11.1 in /src/BERT/training/ots-apple-silicon by @dependabot in #16

New Contributors

Full Changelog: https://github.com/TelecomsXChangeAPi/OpenTextShield/commits/1_7

How to Update

For instructions on upgrading to OTS Version 1.7, please consult our update guide. Should you encounter any issues or require assistance during the update process, our support team is readily available to assist you.

Your feedback and support have been instrumental in the development of OTS. We remain dedicated to improving our platform and encourage you to share your suggestions and inquiries.

Thank you for your ongoing support of OTS!

Contributors

dependabot
Assets 2
Loading