A trigram hidden Markov model for metadata extraction from heterogeneous references

Ojokoh, Bolanle Adefowoke; Zhang, Ming; Tang, Jian

doi:10.1016/j.ins.2011.01.014

Cited by 28 publications

(22 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Yin et al [24] employ a modification of a traditional HMM called a bigram HMM, which considers words' bigram sequential relation and position information. Finally, Ojokoh et al [25] explore a trigram version of HMM, reporting overall accuracy, precision, recall and F1 measure of over 95%.…”

Section: Figure 4: An Example Of a Reference String Represented In Xmentioning

confidence: 99%

Machine Learning vs. Rules and Out-of-the-Box vs. Retrained

Tkaczyk

Collins

Sheridan

et al. 2018

Proceedings of the 18th ACM/IEEE on Joint Conference on Digital Libraries

View full text Add to dashboard Cite

Bibliographic reference parsing refers to extracting machinereadable metadata, such as the names of the authors, the title, or journal name, from bibliographic reference strings. Many approaches to this problem have been proposed so far, including regular expressions, knowledge bases and supervised machine learning. Many open source reference parsers based on various algorithms are also available. In this paper, we apply, evaluate and compare ten reference parsing tools in a specific business use case. The tools are Anystyle-Parser, Biblio, CERMINE, Citation, Citation-Parser, GROBID, ParsCit, PDFSSA4MET, Reference Tagger and Science Parse, and we compare them in both their outof-the-box versions and versions tuned to the project-specific data. According to our evaluation, the best performing out-of-thebox tool is GROBID (F1 0.89), followed by CERMINE (F1 0.83) and ParsCit (F1 0.75). We also found that even though machine learning-based tools and tools based on rules or regular expressions achieve on average similar precision (0.77 for MLbased tools vs. 0.76 for non-ML-based tools), applying machine learning-based tools results in a recall three times higher than in the case of non-ML-based tools (0.66 vs. 0.22). Our study also confirms that tuning the models to the task-specific data results in the increase in the quality. The retrained versions of reference parsers are in all cases better than their out-of-the-box counterparts; for GROBID F1 increased by 3% (0.92 vs. 0.89), for CERMINE by 11% (0.92 vs. 0.83), and for ParsCit by 16% (0.87 vs. 0.75).

show abstract

Section: Figure 4: An Example Of a Reference String Represented In Xmentioning

confidence: 99%

Machine Learning vs. Rules and Out-of-the-Box vs. Retrained

Tkaczyk

Collins

Sheridan

et al. 2018

Proceedings of the 18th ACM/IEEE on Joint Conference on Digital Libraries

View full text Add to dashboard Cite

show abstract

“…Az informatikai módszereket és szövegbányászati eljárásokat tekintve több metódus is szóba jöhet, a szakirodalom alapján az egyik legrelevánsabb megoldás a rejtett Markovmodell alkalmazása (Hetzner, 2008;Ojokoh, Zhang, & Tang, 2011). E módszer mellett a szakemberek más, a mesterséges intelligencia alapján kidolgozott megközelítéseket is alkalmaznak, melyeket általában különféle gépi tanulási algoritmusok segítségével érnek el (Tkaczyk, Bolikowski, Czeczko, & Rusek, 2012;Tkaczyk, Szostek, Fedoryszak, Dendek, & Bolikowski, 2015).…”

Section: Elméleti Háttérunclassified

A Magyar Pedagógia folyóirat tudománymetriai elemzése a hivatkozási szokások és a hivatkozási hálózatok tükrében

Nagy¹,

Molnár²

2018

Magyar Pedagógia

View full text Add to dashboard Cite

“…Two main approaches to reference parsing-related problems are: regular expressions and knowledge-based approaches ( [15], [16]) and machinelearning techniques ( [17], [18], [19]).…”

Section: ) Previous Workmentioning

confidence: 99%

“…Yin et al [18] parse references with the aid of a bigram HMM, in which the emission probability is composed of "beginning" (emitted as first word) and "inner" (inner word) probability. Ojokoh et al [19] propose a full second order HMM with modified Viterbi algorithm and a new smoothing technique for transition probabilities.…”

Section: ) Previous Workmentioning

confidence: 99%

A Modular Metadata Extraction System for Born-Digital Articles

Tkaczyk

Bolikowski

Czeczko

et al. 2012

2012 10th IAPR International Workshop on Document Analysis Systems

View full text Add to dashboard Cite

We present a comprehensive system for extracting metadata from scholarly articles. In our approach the entire document is inspected, including headers and footers of all the pages as well as bibliographic references. The system is based on a modular workflow which allows for evaluation, unit testing and replacement of individual components. The workflow is optimized towards processing of born-digital documents, but may accept scanned document images as well. The machinelearning approaches we have chosen for solving individual tasks increase the ability to adapt to new document layouts and formats. The evaluation tests we have performed showed good results of the individual implementations and the entire metadata extraction process.

show abstract

A trigram hidden Markov model for metadata extraction from heterogeneous references

Cited by 28 publications

References 19 publications

Machine Learning vs. Rules and Out-of-the-Box vs. Retrained

Machine Learning vs. Rules and Out-of-the-Box vs. Retrained

A Magyar Pedagógia folyóirat tudománymetriai elemzése a hivatkozási szokások és a hivatkozási hálózatok tükrében

A Modular Metadata Extraction System for Born-Digital Articles

Contact Info

Product

Resources

About