Matches in ScholarlyData for { ?s ?p This document describes the patents retrieval prototype developed within the MOLTO project. The prototype, accessible at http://molto-patents.ontotext.com, aims to provide an multilingual natural language interface for querying the content of patent documents. We chose the biomedical and pharmaceutical domain for our patent case study.We address the multilingualism of the queries using the GF Tool, which provides an easy way of maintaining a controlled language for the queries. Aiming the best retrieval results of the patent data, the patent documents are pre processed and semantically annotated before storing and indexing them. Furthermore, we store the annotations in the semantic repository OWLIM, which contains a patent specic ontology and other domain specic ontologies. The queries performed in the prototype interface retrieve both the semantic results as RDF instances and a list of patents related to the query. The retrieved patents can be explored in the browser and the user can see in the document the annotations that explain why the particular patent is retrieved. Furthermore, the language of the patent is set by the language of the query. The translation of the patent is provided by the statistical machine translation which is trained on abstracts and claims from an in-domain corpus of patents.. }
Showing items 1 to 1 of
1
with 100 items per page.
- 21 abstract "This document describes the patents retrieval prototype developed within the MOLTO project. The prototype, accessible at http://molto-patents.ontotext.com, aims to provide an multilingual natural language interface for querying the content of patent documents. We chose the biomedical and pharmaceutical domain for our patent case study.We address the multilingualism of the queries using the GF Tool, which provides an easy way of maintaining a controlled language for the queries. Aiming the best retrieval results of the patent data, the patent documents are pre processed and semantically annotated before storing and indexing them. Furthermore, we store the annotations in the semantic repository OWLIM, which contains a patent specic ontology and other domain specic ontologies. The queries performed in the prototype interface retrieve both the semantic results as RDF instances and a list of patents related to the query. The retrieved patents can be explored in the browser and the user can see in the document the annotations that explain why the particular patent is retrieved. Furthermore, the language of the patent is set by the language of the query. The translation of the patent is provided by the statistical machine translation which is trained on abstracts and claims from an in-domain corpus of patents.".