SAMILS : a semi-automatic machine indexer for legal sentences

Abstract

The SAMILS project is set to create an indexed database of judgements given by the Maltese Courts and to allow searching on these judgements. Usually, searching is done manually, which requires a lot of time and resources. This can be eased with the automation of this task. SAMILS was created to provide a set of tools to assist lawyers in building a centralized database of judgments, to allow the indexing of these judgements, and to retrieve the judgements depending on given criteria. Information from the judgements is automatically acquired and stored using XML tags. Indexing is done using a reverse-index structure. A hybrid model between a boolean and statistical model allows easy implementation of the indexing structure whilst allowing some statistical calculations to be done. Since the Maltese language is Semitic, building a lexicon and grammar rules was deemed impossible in the time available. To reduce this shortcoming, a thesaurus is used by the system that allows it to recognize synonyms and relations between words. Conceptual knowledge is stored as keywords in the indexing structure. The thesaurus is one of the keys, which links these concepts together. An Internet-based homepage allows users to use the system from the comfort of their home or office.