Each year, hundreds of public works projeccts are performed. Each project must be examined and assessed on their potential environmental impact. Major projects are discussed in LegCo's Public Works Subcommittee (PWSC) meetings with environmental assessment performed by the Environmental Protection Department (EPD).
To suppport EPD's work, we have created a text-mining/information retrieval system that automatically analyzes PWSC meeting papers to retrieve related content from historical records. These historical records may be previous PWSC meeting minutes, PWSC papers, EPD briefing notes, etc. Our system acts as a research tool for EPD officers. Results from our system is used to create briefing notes to support discussions in the LegCo's PWSC meetings.
Our client is the Hong Kong Environmental Protection Department (EPD). The text-mining/information retrieval system supports their work in LegCo's Public Works Subcommittee (PWSC) meetings.
The Legislative Council of Hong Kong (LegCo) is the legislature of the Hong Kong Special Administrative Region (SAR). The Finance Committee is one of three standing committees under the Council. There are 2 subcommittees under the Finance Committee: the Establishment Subcommittee and the Public Works Subcommittee (PWSC). The AI text-mining work was performed to support discussions in the PWSC meetings.
The Public Works Subcommittee examines and makes recommendations to the Finance Committee on the Government's expenditure proposals under the Capital Works Reserve Fund for projects in the public works programme and building projects carried out by or on behalf of subvented organizations. This committee meets roughly once or twice each month. The Hong Kong Environmental Protection Department (EPD) is a regular member of the PWSC meetings.
Any PWSC paper to be discussed in the PWSC meetings that relates to the environment is examined by EPD. Prior to the PWSC meetings, EPD officers prepare meeting briefing notes (BN) that contains environmental protection information. References to related environmental discussions in previous meetings are highlighted as well. In the past, research to suppport the drafting of briefing notes is done manually. This is very time-consuming and error prone. Our text mining system automatically performs this research within seconds and error free.
The AI technology we used included search engine, text mining techniques, keyword analysis using a dictionary of environmental protection-related words. The system supports both English and Chinese.
The main objective of the project is to streamline EPD's research and to improve efficiency and productivity. In the past, the quality of PWSC briefing notes depends heavily on the experience and memory of an individual officer. With our text mining system, nothing falls through the cracks. All relevant historical cases, past discussions, and decisions are retrieved automatically.