Machine Comprehension of Legal Text

Principal Investigator: Professor Benjamin C.M. KAO (Professor from Department of Computer Science)

This project is showcased in the second exhibition – Digitization in Innovation Wing Two

About the scholar

Professor Benjamin C.M. KAO

Research Interests:
• Database management systems
• Data mining
• Real-time systems
• Information retrieval systems

Project information
The Law and Technology Centre, co-directed by Professor Kao, aims to advance research and provide public service in the interdisciplinary area of information technology and law. The current research focuses on the integration of artificial intelligence (AI) into legal research and practice.

Awarded and funded projects include:

AI Hong Kong Legal Information Institute (AI-HKLII)
HKLII provides public access to legal information including judgments and legislation. We develop latest AI technologies to achieve:

  • Statistical query engine: By converting plaintext judgments into structured records, our system allows users to perform sophisticated data analytics.
  • Judgment synopsis composer: A synopsis provides a textual summary of a judgment and displays the salient feature values to users.
  • Semantic search engine and document recommender: We employ advanced techniques to analyze the semantics of the user queries to generate better search and recommendation results.

AI Community Legal Information Centre (AI-CLIC)

  • The CLIC website introduces commonly enquired legal questions by providing contents that can be easily understood by the non-expert.
  • We simplify CLIC’s information search by introducing a dialogue module which comprehends natural language text or speech and interacts with users via dialogues.
  • When a user provides a description in the form of a brief story, the dialogue module would predict the user’s search intents and recommend relevant contents to the user.

Sentencing Predictor for Drug trafficking

  • The sentencing predictor uses the power of AI to predict sentences for the offence of drug trafficking with explanations for 8 types of commonly used dangerous drugs. It also provides some relevant court decisions.
  • The sentencing predictor helps lawyers and social workers to make confident predictions and reduce research time and cost. It also serves to educate the public of the likely legal consequences of committing drug trafficking offences.
Project images
Content page of AI-HKLII: The new HKLII interface provides accesses to AI functionalities, such as the document recommender.
Textual Summary of a Judgment: The most important parts of the judgment are highlighted and displayed to the users.
Statistical Query Engine: The engine is capable of handling analytical queries submitted by the users.
Dialog System on AI-CLIC: After a user describes the situation, the system will recommend the most relevant CLIC pages to the users.
Sentencing Predictor for Drug Trafficking: The predictor can accurately predict the sentence by asking the user some simple questions.
Achievement of the Project
Enquiry / Feedback

Please feel free to give your enquiry / feedbacks to the research team by filling the form ( Thank you!