Mining MEDLINE: Abstracts, Sentences, or Phrases?

By Jing Ding, D. Berleant, Dan Nettleton, E. Wurtele. Pacific Symposium on Biocomputing, 2001

297 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Molecular Biology

View Paper: https://www.semanticscholar.org/paper/a846c4d28a2d5986ba86b1ec8a6dc3b67d3aa0f3

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Molecular Biology

Citations in the corpus

Transfer learning for biomedical named entity recognition with neural networks (2018, background): “…(i.e., GSCs), each containing hand-labeled annotations for one of these entity types, such as the “CDR” corpus for chemicals (Li et al., 2016), “NCBI Disease” for disease names (Doğan et al., 2014), “S800” for species (Pafilis et al., 2013) and “DECA” for genes/proteins (Wang et al., 2010).”
Extracting interactions between proteins from the literature (2008, result): “[15] compared the results of employing different text units such as phrases, sentences, and abstracts from MEDLINE to mine interactions between biochemical entities based on co-occurrences.”
Literature mining for the biologist: from information retrieval to biological discovery (2006, methodology): “Co-occurrence methods tend to give better recall but worse precision than NLP method”
A Comprehensive Benchmark of Kernel Methods to Extract Protein–Protein Interactions from Literature (2010, methodology): “In [36], a thorough analysis of five freely available PPIannotated resources, namely AIMed [18], BioInfer [59], HPRD50 [60], IEPA [61], and LLL [62], was performed.”
Mining the Biomedical Literature in the Genomic Era: An Overview (2003, unknown): “Moreover, since these methods depend on the co-occurrence of terms, within a sentence, a phrase, or an abstract (see Ding et al. [2002] for a discussion on granularity choice), they can only reveal relationships that are already reported in the literature and do not attempt to detect new relations.”
Deep learning with word embeddings improves biomedical named entity recognition (2017, methodology): “, 2003; Eltyeb and Salim, 2014). These patterns vary depending on the specific textual properties of an entity class. The definition of such entityspecific patterns is time consuming and requires domain-expert knowledge. Dictionary-based methods extract named entities by searching them in dictionari”
Bridging semantics and syntax with graph algorithms - state-of-the-art of extracting biomedical relations (2017, methodology): “Shared resources for biomedical relation extraction Utility category Data sources Terminology/Ontology/Knowledge Base GO [161], UMLS [58], MeSH [162], Wordnet [118], Verbnet [163], Biothesaurus [122], PDB [101], Uniprot [121], SIDER [164], DrugBank [165], HIVDB [166], RegaDB [167], Entrez Gene [153]”
Content-rich biological network constructed by mining PubMed abstracts (2004, background): “The texts (including each title and abstract) are then parsed into units of one sentence, which has been shown to yield higher performance levels than paragraphs or phrases in the identification of relationships from MEDLINE abstracts [19].”
All-paths graph kernel for protein-protein interaction extraction with evaluation of cross-corpus learning (2008, methodology): “We evaluate our method using five publicly available corpora that contain PPI interaction annotation: AImed [4], BioInfer [6], HPRD50 [19], IEPA [20] and LLL [21].”
DISEASES: Text mining and data integration of disease–gene associations (2014, background): “We focus on the latter approach, which is highly flexible and generally gives better recall, but worse precision, than NLP[18–20].”