Evaluation Methods for Question Decomposition Formalism

Computational Linguistics and Natural Language Processing, May 2024 in Singapore

Introducing a paper about two methods for the evaluation of Question Decomposition Meaning Representation (QDMR) as predicted by sequence-to-sequence model and COPYNET parser for natural language questions processing, motivated by the fact that previous evaluation metrics used for this task do not take into account some characteristics of the representation, such as partial ordering structure. To this end, several heuristics to extract such partial dependencies are formulated, followed by the hereby proposed evaluation methods denoted as Proportional Graph Matcher (PGM) and Conversion to Normal String Representation (Nor-Str), designed to better capture the accuracy level of QDMR predictions. Experiments are conducted to demonstrate the efficacy of the proposed evaluation methods and show the added value suggested by one of them- the Nor-Str, for better distinguishing between high and low-quality QDMR when predicted by models such as COPYNET. This work represents an important step forward in the development of better evaluation methods for QDMR predictions, which will be critical for improving the accuracy and reliability of natural language question-answering systems.

Promt Engineering Techniques & Tricks

In this session you'll meet Aviv, senior development engineer from Microsoft, who shall demonstrate techniques for prompt engineering used to crack the renowned Gandalf AI challenge (where only 8% completed level 7).
The session will cover the techniques used and architectures of diffrent riddles in that challenge.

Ask the Expert - Group B (Thu 16:50 to 17:40)

'Got a problem? Don't know who to call? Bring it to SQLBits and get it solved by our panel of experts!

More information can be found here:
https://sqlbits.com/news/ask-the-experts-at-the-experts-lounge/

5. Data Science and Analytics: Lewis Prince
7. Programming and Development: Aviv Yaniv
8. Data Management and Governance: Ust Oldfield
9. User Experience and Design:

Scale your business in real-time using ML in Microsoft Fabric Real-time Analytics

When flights are delayed or cancelled, airport and airline operations face a huge challenge: how to serve the thousands of affected passengers quickly and efficiently. You may have experienced the frustration of long queues, insufficient staff, lack of refreshments, or missing baggage. But what if there is a better way to handle this situation with the power of machine learning?
In this session, you will learn how Microsoft Fabric Real-time Analytics uses machine learning to solve the autoscaling problem for customer service. You will see how an ML model can:
• Predict the number of additional staff needed based on real-time data
• Prioritize and queue passengers based on their PNR information
• Optimize the allocation of resources and services
This session is for anyone who is interested in how machine learning can improve customer satisfaction and operational efficiency. You will get to see Senior Software Engineer showing code, queries, and ML models behind the scenes, as well as a live demonstration of the autoscaling in action.
Join us for this exciting technical session to discover how Microsoft Fabric Real-time Analytics can help you scale up your customer service with machine learning in real-time.