A resource-saving collective approach to biomedical semantic role labeling

Abstract

BackgroundBiomedical semantic role labeling (BioSRL) is a natural language processing technique that identifies the semantic roles of the words or phrases in sentences describing biological processes and expresses them as predicate-argument structures (PAS’s). Currently, a major problem of BioSRL is that most systems label every node in a full parse tree independently; however, some nodes always exhibit dependency. In general SRL, collective approaches based on the Markov logic network (MLN) model have been successful in dealing with this problem. However, in BioSRL such an approach has not been attempted because it would require more training data to recognize the more specialized and diverse terms found in biomedical literature, increasing training time and computational complexity.ResultsWe first constructed a collective BioSRL system based on MLN. This system, called collective BIOSMILE (CBIOSMILE), is trained on the BioProp corpus. To reduce the resources used in BioSRL training, we employ a tree-pruning filter to remove unlikely nodes from the parse tree and four argument candidate identifiers to retain candidate nodes in the tree. Nodes not recognized by any candidate identifier are discarded. The pruned annotated parse trees are used to train a resource-saving MLN-based system, which is referred to as resource-saving collective BIOSMILE (RCBIOSMILE). Our experimental results show that our proposed CBIOSMILE system outperforms BIOSMILE, which is the top BioSRL system. Furthermore, our proposed RCBIOSMILE maintains the same level of accuracy as CBIOSMILE using 92% less memory and 57% less training time.ConclusionsThis greatly improved efficiency makes RCBIOSMILE potentially suitable for training on much larger BioSRL corpora over more biomedical domains. Compared to real-world biomedical corpora, BioProp is relatively small, containing only 445 MEDLINE abstracts and 30 event triggers. It is not large enough for practical applications, such as pathway construction. We consider it of primary importance to pursue SRL training on large corpora in the future.

BelSmile: a biomedical semantic role labeling approach for extracting biological expression language from text

“…BioSRL is usually formulated as a supervised machine learning problem that relies on manually annotated training corpora ( 4 , 13 ). However, building such large corpora requires much human effort.…”

“…Third, function patterns are used to determine the functions of the NEs. Finally, the SRL-based method classifies ( 4 ) the causal and correlative relationships.…”

“…It can be represented by a PAS in which ‘reduced’ is the predicate, ‘Inhibition of COX2’ and ‘both IL-1 beta and IL-6 release’ comprise ARG0 (agent) and ARG1 (patient), respectively, with ‘markedly’ as the ARGM-MNR (manner). The SRL component used in our system consists of two components, RCBiosmile ( 4 ) and a syntactic-based labeler (SBL).

Figure 3.

…”

“…Subsequently, BelSmile uses a rule-based method to normalize NEs and classify their functions. Lastly, BelSmile uses a revised RCBiosmile ( 4 ), to extract subject–verb–object (SVO) tuples and determine the relation type. Each component of BelSmile will be explained in the following section.…”

The extraction of complex relationships and their conversion to biological expression language (BEL) overview of the BioCreative VI (2017) BEL track

“…In the SRL stage, two systems, RCBiosmile (22) and Enju (23), are employed. Enju covers some predicates (verbs) not recognized by RCBiosmile, which is trained on the BioProp corpus (24).…”