Modeling the Right to Information Query Log: Learning Latent Parameters to Identify Amendment Scopes in the RTI Act
No Thumbnail Available
Amendments to laws are necessary to keep up with the changing needs of the society. Such a process is largely manual, and policy-makers take feedback from the society for the introduction of an amendment. These include statistics and other observations taken into account by law makers during the amendment process. The Right to information (RTI) Act 2005 gives Indian citizens the opportunity to interact with the government. Any Indian citizen can question public authorities and demand information from them. Evidence suggests that the RTI query-reply process has also sown seeds towards the proposal as well as the introduction of amendments to the RTI Act. This direct interaction between citizens and the government have latent information that leads towards potential amendment scopes. The presence of such pointers encourages us to think that tentative amendments can be proposed by a learning process, and that feedback for potential amendments can be identified by an extensive analysis of the database of RTI queries and their reply-statistics. The objective of this thesis is to analyze RTI query and reply statistics data to mine latent patterns that act as feedback for potential amendments to the existing RTI laws. We have collected RTI applications containing queries of citizens. From this, relevant RTI properties are identified and their definitions and indicators are studied from the literature. These definitions are used for proposing data models, and learning models are used to quantify RTI parameters. The quantified RTI parameters obtained in the course of this thesis reveal patterns that suggest several scopes for potential amendments in the RTI laws.
Supervisor: V. Vijaya Saradhi
COMPUTER SCIENCE AND ENGINEERING