Exploiting semantic annotations for open information extraction: an experience in the biomedical domain

Abstract

Abstract.The increasing amount of unstructured text published on the Web is demanding new tools and methods to automatically process and extract relevant information. Traditional information extraction has focused on harvesting domain-specific, pre-specified relations, which usually requires manual labor and heavy machinery. Especially in the biomedical domain the main efforts have been directed towards the recognition of welldefined entities such as genes or proteins, which constitutes the basis for extracting the relations between the recognized entities. The intrinsic features and scale of the Web demand new approaches able to cope with the diversity of documents, where the number of relations is unbounded and not known in advance. This paper presents a scalable method for the extraction of domain-independent relations from text that exploits the knowledge in the semantic annotations. The method is not geared to any specific domain (e.g., protein-protein interactions, drug-drug interactions, etc.) and does not require any manual input or deep processing. Moreover, the method uses the extracted relations to compute groups of abstract semantic relations characterized by their signature types and synonymous relation strings. This constitutes a valuable source of knowledge when constructing formal knowledge bases, as we enable seamless integration of the extracted relations with the available knowledge resources through the process of semantic annotation. The proposed approach has successfully been applied to a large text collection in the biomedical domain and the results are very encouraging.

Generating disease-pertinent treatment vocabularies from MEDLINE citations

“…Based on the error analysis in Section 4.4, approximately two thirds of the false negative concepts were probably attributed to the relation extraction tool we have used. However, there exist many other approaches aimed at extracting semantic relations from the biomedical literature or web documents, and some of them were also used UMLS and/or MetaMap [43]. Therefore, our system may gain further recall by incorporating the output of other relation extraction approaches or tools as secondary knowledge sources in addition to the SemMedDB to our proposed pipeline process.…”

Open Information Extraction with Meta-pattern Discovery in Biomedical Literature

“…For example, Attias et al [28] present an OpenIE system for biomedical named entity recognition based on NELL [27], and propose a method for assessing seed qualities to prevent semantic drift. Nebot et al [30,31] propose a scalable method to extract surfaceform biomedical relationships not specific to any relation type and Input: raw text Amphetamine and cocaine decreased susceptibility to myoclonus in young mice and increased susceptibility in mature mice. <CHEMICAL>Amphetamine</ CHEMICAL> and <CHEMICAL>cocaine</CHEMICAL> decreased susceptibility to <DISEASE>myoclonus</DISEASE> in young <SPECIES>mice</ SPECIES> and increased susceptibility in mature <SPECIES>mice</SPECIES>.…”

Semantic Annotation of Unstructured Documents Using Concepts Similarity

“…Nebot and Berlanga [27] explore the use of semantic annotation in the biomedical domain. They present a scalable method to extract domain-independent relationships.…”