An integrated architecture for shallow and deep processing

Abstract

We present a flexible architecture for the integration of shallow and deep NLP components which is aimed at flexible combination of different language technologies for a range of practical current and future applications. In particular, we describe the integration of a high-level HPSG parsing system with different high-performance shallow components, ranging from named entity recognition to chunk parsing and shallow clause recognition. The NLP components enrich a representation of natural language text with layers of new XML meta-information using a single shared data structure, called the text chart. We describe details of the integration methods, and show how information extraction and language checking applications for real-world German text benefit from a deep grammatical analysis. An Integrated Architecture for Shallow and Deep ProcessingPaper ID: P0270 AbstractWe present a flexible architecture for the integration of shallow and deep NLP components which is aimed at flexible combination of different language technologies for a range of practical current and future applications.In particular, we describe the integration of a high-level HPSG parsing system with different high-performance shallow components, ranging from named entity recognition to chunk parsing and shallow clause recognition. The NLP components enrich a representation of natural language text with layers of new XML meta-information using a single shared data structure, called the text chart. We describe details of the integration methods, and show how information extraction and language checking applications for real-world German text benefit from a deep grammatical analysis.

An approach towards the reconstruction of regulatory networks

“…The main goal of this step is to use an approach to identify these interactions by using some Natural Language Processing (NLP) method [Spyns 1996] such as Shallow and Deep processing [Crysmann et al 2002] or Dependency parsing [Kubler et al 2009]. …”

Ontology-based information extraction and integration from heterogeneous data sources

“…A number of methods have been proposed for the integration of ''deep'' and ''shallow'' grammar processing models in so-called hybrid NLP architectures, which try to combine the robustness of shallow processing tools with the higher precision and finegrainedness of deep linguistic analysis (cf. Crysmann et al, 2002;Frank et al, 2003Frank et al, , 2004.…”

“…A number of methods have been proposed for the integration of ''deep'' and ''shallow'' grammar processing models in so-called hybrid NLP architectures that combine the robustness of shallow processing tools with the higher precision and finegrainedness of deep linguistic analysis (cf. Crysmann et al, 2002;Frank et al, 2003Frank et al, , 2004. While these earlier proposals tried to integrate ''lower-level'' information provided by robust shallow processing tools to improve the robustness and coverage of ''deeper'' analysis systems, the architecture realized in SOBA is designed to work in the opposite direction.…”

A Hybrid Method for Open Information Extraction Based on Shallow and Deep Linguistic Analysis

“…For this purpose, different threshold values of confidence and sentence length are applied and the impact of these factors on resulting performance is assessed. Similar hybrid techniques have also been employed in other fields such as parsing (e.g., in [26]), where an architecture for integration of shallow and deep NLP components have been presented. In current work, since our goal is to improve performance in a fast way, DepOE [22] system is used.…”