本文主要介绍了生物学文本挖掘的研究现状、待解决的问题、应用前景以及现有可利用的文本挖掘资源。从传统的纸质学术刊物到网络信息共享，由于信息资源更加容易访问，生物学家可以更方便地获得大量文献。但是由于文献数量之多，研究人员要找到所需的信息越来越困难。因此，最近的生物学竞赛侧重于评价各种搜索文献的方法。在这些方法中，文本挖掘技术显示了极大的优势。随着近年来文本挖掘技术的发展，同时出版商用XML格式存储文献，越来越多的文本挖掘系统（TMS）可用于文献管理，保持资料完整性，并确保可以适当地链接到其它资源以获得资料数据。即便如此，一些新的挑战已经出现在全文分析、生命科学术语、复杂关系抽取和信息融合。为了使文本挖掘更为有效，必须克服这些问题。在本文中，作者明确了文本挖掘中遇到的挑战，探讨如何克服这些问题，以及考虑应用现有可用资源帮助我们达到预期目标。
文章首先介绍了文本挖掘的研究背景和研究意义。由于文本挖掘使得所需的生物学文献更容易获取，它在各种方法中显示出了很大的优势，更为有效。比较而言，信息检索的目的是帮助用户寻找满足他们需求的信息，用户用一组关键词来表达检索意图，系统返回给用户一组包含这些关键词的文档。在信息检索中没有“产生”新的信息，而文本挖掘的目的是从文本数据中发现和产生新的知识。它将文本集合看作是知识库，而不是简单地看作是多个文本的集合。它检查文本集合中的每一个文本，抽取文本信息，然后提出关于新知识的假设，并验证它。
接下来文章提出了生物学文本挖掘主要面临的挑战，主要有以下三点：一，专业术语识别，例如基因名称和对应的数据库标示符如此之多，即使是专家要理解这些术语并保持更新和修改也是比较困难的；二，全文分析需要应用更为复杂的自然语言处理（NLP），这些不是当前生物信息检索和提取工具所能胜任的；三，当前的文本挖掘系统（TMS）无法整合来自不同资源、具有不同上下文和不同排版格式的信息。文本挖掘处理技术中主要包括：命名实体识别和跨物种规范化、关联关系和事实的提取（相关与无关的信息处理和规则提取）。
在文本挖掘的应用前景中，主要提出了以用户为中心的设计思想、生物信息学研究人员之间的沟通与协作、以及信息的融合。文本挖掘的研究人员通常善于分析文本内容，但他们并不善于设计对于普通用户来说很容易上手的交互系统。要解决此问题，研究人员必须设计具有直观界面的、需要很少或根本不需要文本挖掘背景知识和NLP技术的应用。这样做的目的是为生物信息学，生物学，生物医学和药理学的研究人员提供对于生物性反应的高层次的理解，并帮助他们形成新的假设。
最后文章给出了研究中可以利用的文本挖掘资源，例如特定领域的词汇，词典，术语标准，本体，和其他评价的基于任务的挑战是非常重要的。这些资源主要有：通过基于任务目标的竞赛来评估和促进文本挖掘研究，以及可用于研究的公共文本挖掘语料库。其中文本挖掘语料库包含命名实体识别语料库、关系提取语料库、词性语法语义标注和全文语料库。希望通过这些现有的资源可以更好地促进文本挖掘这一领域的进一步发展。
New Challenges for Biological Text-Mining in the Next Decade
Hong-Jie Dai1,2, Yen-Ching Chang1, Richard Tzong-Han Tsai3, and Wen-Lian Hsu1,2, Fellow, IEEE
1Institute of Information Science, "Academia Sinica'', 115, Taiwan, China
2Department of Computer Science, "National Tsing-Hua University'', 300, Taiwan, China
3Department of Computer Science and Engineering, Yuan Ze University, 320, Taiwan, China
The massive flow of scholarly publications from traditional paper journals to online outlets has benefited biologists because of its ease to access. However, due to the sheer volume of available biological literature, researchers are finding it increasingly difficult to locate needed information. As a result, recent biology contests, notably JNLPBA and BioCreAtIvE, have focused on evaluating various methods in which the literature may be navigated. Among these methods, text-mining technology has shown the most promise. With recent advances in text-mining technology and the fact that publishers are now making the full texts of articles available in XML format, TMSs can be adapted to accelerate literature curation, maintain the integrity of information, and ensure proper linkage of data to other resources. Even so, several new challenges have emerged in relation to full text analysis, life-science terminology, complex relation extraction, and information fusion. These challenges must be overcome in order for text-mining to be more effective. In this paper, we identify the challenges, discuss how they might be overcome, and consider the resources that may be helpful in achieving that goal.
This work was supported by the "National Science Council'' under Grant Nos. NSC 97-2218-E-155-001 and NSC96-2752-E-001-001-PAE, the Research Center for Humanities and Social Sciences, and the Thematic Program of "Academia Sinica'' under Grant No. AS95ASIA02.
Hong-Jie Dai received his B.S. degree in computer science and information engineering from Tung Hai University and his M.S. degree in computer science and information engineering from "National Central University in Taiwan'' in 2003 and 2005, respectively. Since 2005, he has been an assistant researcher at the "Academia Sinica''. His research interests include: bioinformatics, machine learning, text mining, natural language processing and software engineering.
Yen-Ching Chang received her B.S. degree in biochemical science and technology from "National Taiwan University'' and her M.S. degree in biochemistry and molecular biology from "National Taiwan University'', College of Medicine. Since 2008, she has been an research assistant at the "Academia Sinica''. Her research interests include proteomics and text mining.
Richard Tzong-Han Tsai received the B.S. degree in computer science and information engineering from "National Taiwan University'', Taipei, in 1997, the M.S. degree in computer science and information engineering from "National Taiwan University'' in 1999, and the Ph.D. degree in computer science and information engineering from "National Taiwan University'' in 2006. He was a postdoctoral fellow at "Academia Sinica'' from 2006 to 2007. He is now an assistant professor of Department of Computer Science and Engineering, Yuan Ze University, Zhongli, Taiwan, China. His research areas are natural language processing, cross-language information retrieval, biomedical literature mining, and information services on mobile devices.
Wen-Lian Hsu is a distinguished research fellow in the Institute of Information Science, "Academia Sinica''. He received his B.S. degree from the Department of Mathematics, "National Taiwan University'' in 1973 and his Ph.D. degree in operations research from Cornell University in 1980, respectively. He then joined Northwestern University, and was promoted to tenured associate professor in 1986. In 1989, he joined the Institute of Information Science as a research fellow. Earlier in his career, Dr. Hsu focused on theoretical graph algorithms and frequently published papers in top-notch journals, such as JACM, SIAM Journal on Computing. After returning to Taiwan, China, he started research on automatic conversion of Pinyin to characters. In 1993, he invented a Chinese natural input method which has since attracted two million users and revolutionized the phonetic input for Chinese in Taiwan, China. Later, he moved into question answering, and bioinformatics. He is currently the director of the International Graduate Program in Bioinformatics in "Academia Sinica''. Dr. Hsu received many awards including the Distinguished Research Fellow Award of the "National Science Council'', K. T. Li breakthrough award, IEEE fellow, "Academia Sinica Investigator'' Award, and Teco Technology Award. From 2001 to 2002, he was the President of the Artificial Intelligence Society in Taiwan, China.
[1] Kim J D et al. Introduction to the bio-entity recognition task at JNLPBA. In Proc. the International Workshop on Natural Language Processing in Biomedicine and Its Applications (JNLPBA2004), Geneva, Switzerland, Aug. 28-29, 2004, pp.70-75.
[2] Hirschman L et al. Overview of BioCreAtIvE: Critical assessment of information extraction for biology. BMC Bioinformatics, 2005, 6(Suppl.1): S1.
[3] Krallinger M et al. Evaluation of text-mining systems for biology: Overview of the Second BioCreative community challenge. Genome Biology, 2008, 9(Suppl. 2): S1.
[4] Hearst M A. Untangling text data mining. In Proc. the 37th Annual Meeting of the Association for Computational Linguistics on Computational Linguistics, College Park, USA, June 20-26, 1999, pp.3-10.
[5] Hahn U et al. Text mining: Powering the database revolution. Nature, 2007, 448(7150): 130.
[6] HearstM.What is text mining. 2003, http://people.ischool.berkeley. edu/∼hearst/text-mining.html.
[7] Dai H J et al. BIOSMILE web search: A web application for annotating biomedical entities and relations. Nucl. Acids Res., 2008, 36(Web Sever Issue): W390-W398.
[8] Rebholz-Schuhmann D et al. Text processing through Web services: Calling Whatizit. Bioinformatics, 2008, 24(2): 296- 298.
[9] Fern′andez J M et al. iHOP web services. Nucl. Acids Res., 2007, 35(Web Server Issue): W21-W26.
[10] Elsevier Article 2.0 Contest. http://article20.elsevier.com/contest/ home.html, Accessed July, 2009.
[11] The Elsevier Grand Challenge. http://www.elseviergrandchallenge. com/, Accessed November, 2009.
[12] BioCreAtIvE II.5. http://www.biocreative.org/events/biocreative- ii5/biocreative-ii5/, Accessed December, 2009.
[13] Ananiadou S, Chruszcz J et al. The national ventre for text mining: Aims and objectives. In Proc. UKKDD2007, Kent, UK, April 25, 2007, pp.6-12.
[14] RSC Project Prospect. http://www.projectprospect.org/.
[15] Seringhaus M, Gerstein M. Manually structured digital abstracts: A scaffold for automatic text mining. FEBS Letters, 2008, 582(8): 1170.
[16] Morgan A et al. Overview of BioCreative II gene normalization. Genome Biology, 2008, 9(Suppl. 2): S3.
[17] Gonzalez G et al. Mining gene-disease relationships from biomedical literature: Weighting protein-protein interactions and connectivity measures. In Proc. the Pacific Symposium on Biocomputing, 2007, 12: 28-29.
[18] Tsai R T H, Lai P et al. HypertenGene: Extracting key hypertension genes from biomedical literature with position and automatically-generated template features. BMC Bioinformatics, 2009, 10(Suppl. 5): S9.
[19] Cohen A M, Hersh W R. A survey of current work in biomedical text mining. Briefings in Bioinformatics, 2005, 6(1): 57- 71.
[20] Smith L et al. Overview of BioCreative II gene mention recognition. Genome Biology, 2008, 9(Suppl.2): S2.
[21] Krallinger M et al. Overview of the protein-protein interaction annotation extraction task of BioCreative II. Genome Biology, 2008, 9(Suppl.2): S4.
[22] Chinchor N. MUC-7 named entity task definition (Version 3.5). In Proc. the 7th Message Understanding Conference, 1997.
[23] Leser U, Hakenberg J. What makes a gene name? Named entity recognition in the biomedical literature. Briefings in Bioinformatics, 2005, 6(4): 357-369.
[24] Erhardt R A A et al. Status of text-mining techniques applied to biomedical text. Drug Discovery Today, 2006, 11(7/8): 315-325.
[25] Liu H et al. A study of abbreviations in MEDLINE abstracts. In Proc. AMIA Annual Symposium, San Antonio, USA, Nov. 9-13, 2002, pp.464-468.
[26] Tanabe L, Wilbur W J. Tagging gene and protein names in full text articles. In Proc. the ACL-02 Workshop on Natural Language Processing in the Biomedical Domain—Volume 3, Philadelphia, USA, July 11, 2002, pp.9-13.
[27] Tanabe L, Wilbur W J. Tagging gene and protein names in biomedical text. Bioinformatics, 2002, 18(8): 1124-1132.
[28] Zhao S. Named entity recognition in biomedical texts using an HMM model. In Proc. the International Joint Workshop on Natural Language Processing in Biomedicine and Its Applications, Geneva, Switzerland, Aug. 28-29, 2004, pp.84-87.
[29] Kazama J i et al. Tuning support vector machines for biomedical named entity recognition. In Proc. the ACL-02 Workshop on Natural Language Processing in the Biomedical Domain — Volume 3, Philadelphia, USA, July 11, 2002, pp.1-8.
[30] Finkel J et al. Exploiting context for biomedical entity recognition: From syntax to the web. In Proc. the International Joint Workshop on Natural Language Processing in Biomedicine and Its Applications, Geneva, Switzerland, Aug. 28-29, 2004, pp.88-91.
[31] Tsai R T H et al. NERBio: Using selected word conjunctions, term normalization, and global patterns to improve biomedical named entity recognition. BMC Bioinformatics, 2006, 7(Suppl. 5): S11.
[32] Si L et al. Boosting performance of bio-entity recognition by combining results from multiple systems. In Proc. the 5th International Workshop on Bioinformatics, Chicago, USA, Aug. 21, 2005, pp.76-83.
[33] Altman R et al. Text mining for biology — The way forward: Opinions from leading scientists. Genome Biology, 2008, 9(Suppl. 2): S7.
[34] Jimeno A et al. Assessment of disease named entity recognition on a corpus of annotated sentences. BMC Bioinformatics, 2008, 9(Suppl. 3): S3.
[35] Yu H et al. Mapping abbreviations to full forms in biomedical articles. Journal of the American Medical Informatics Association, 2002, 9(3): 262-272.
[36] Schwartz A S, Hearst M A. A simple algorithm for identifying abbreviation definitions in biomedical text. Proc. Pac. Symp. Biocomput., 2003, 8: 451-462.
[37] Podowski R et al. Suregene, a scalable system for automated term disambiguation of gene and protein names. Journal of Bioinformatics and Computational Biology, 2005, 3(3): 743- 770.
[38] Hirschman L et al. Overview of BioCreAtIvE task 1B: Normalized gene lists. BMC Bioinformatics, 2005, 6(Suppl. 1): S11.
[39] Cohen W, Minkov E. A graph-search framework for associating gene identifiers with documents. BMC Bioinformatics, 2006, 7: 440.
[40] Leitner F. Comparative community assessments for applied biomedical text mining: BioCreative II challenge and metaservices. In Intelligent Systems for Molecular Biology (ISMB) and European Conference on Computational Biology (ECCB), Highlights Track, Stockholm, Sweden, June 27-July 2, 2009.
[41] Fundel K, Guttler D et al. A simple approach for protein name identification: Prospects and limits. BMC Bioinformatics, 2005, 6(Suppl. 1): S15.
[42] Hakenberg J et al. Me and my friends: Gene mention normalization with background knowledge. In Proc. the Second BioCreAtIvE Challenge Evaluation Workshop, Madrid, Spain, April 23-25, 2007, p.23-25.
[43] Seki K, Javed M. Discovering implicit associations between genes and hereditary diseases. In Proc. Pac. Symp. Biocomput., 2007, 12: 316-327.
[44] Cooper J W, Kershenbaum A. Discovery of protein-protein interactions using a combination of linguistic, statistical and graphical information. BMC Bioinformatics, 2005, 6: 143.
[45] Shah P K et al. Information extraction from full text scientific articles: Where are the keywords? BMC Bioinformatics, 2003, 4: 20.
[46] Shatkay H et al. Integrating image data into biomedical text categorization. Bioinformatics, July 15, 2006, 22(14): e446- e453.
[47] Kou Z et al. A stacked graphical model for associating information from text and images in figures. In Proc. Pac. Symp. Biocomput., 2007, 12: 257-268.
[48] Saric J et al. Extraction of regulatory gene/protein networks from Medline. Bioinformatics, March 15, 2006, 22(6): 645- 650.
[49] Ono T et al. Automated extraction of information on proteinprotein interactions from the biological literature. Bioinformatics, Feb. 2001, 17(2): 155-161.
[50] Kim S et al. Kernel approaches for genic interaction extraction. Bioinformatics, 2008, 24(1): 118-126.
[51] Bunescu R, Mooney R. Subsequence kernels for relation extraction. Advances in Neural Information Processing Systems, 2006, 18: 171-178.
[52] Barnickel T et al. Large scale application of neural network based semantic role labeling for automated relation extraction from biomedical texts. PLoS One, 2009, 4(7): e6393.
[53] Ramani A et al. Consolidating the set of known human protein-protein interactions in preparation for large-scale mapping of the human interactome. Genome Biology, 2005, 6(5): R40.
[54] Bunescu R et al. Comparative experiments on learning information extractors for proteins and their interactions. Artificial Intelligence in Medicine, 2005, 33(2): 139-155.
[55] Rosario B, Hearst M A. Multi-way relation classification: Application to protein-protein interactions. In Proc. the Conference on Human Language Technology and Empirical Methods in Natural Language Processing, Vancouver, Canada, Oct. 6- 8, 2005, pp.732-739.
[56] Craven M, Kumlien J. Constructing biological knowledge bases by extracting information from text sources. In Proc. the 7th International Conference on Intelligent Systems for Molecular Biology, Heidelberg, Germany, Aug. 6-10, 1999, pp.77-86.
[57] Rindflesch T C et al. EDGAR: Extraction of drugs, genes and relations from the biomedical literature. In Proc. Pac. Symp. Biocomput., 2000, 5: 514-525.
[58] Chun H W et al. Extraction of gene-disease relations from Medline using domain dictionaries and machine learning. In Proc. the Pacific Symposium on Biocomputing, 2006, 11: 4- 15.
[59] Tsai R T H et al. HypertenGene: Extracting key hypertension genes from biomedical literature with position and automatically-generated template features. To appear in BMC Bioinformatics, 2009.
[60] Miyao Y, Sagae K et al. Evaluating contributions of natural language parsers to protein-protein interaction extraction. Bioinformatics, 2008, 25(3): 394-400.
[61] Wong L. PIES, a protein interaction extraction system. In Proc. Pacific Symposium on Biocomputing, 2001, 6: 520-531.
[62] Casta˜no J et al. Anaphora resolution in biomedical literature. In International Symposium on Reference Resolution for NLP, Alicante, Spain, June 3-4, 2002.
[63] Pustejovsky J et al. Medstract: Creating large-scale information servers for biomedical libraries. In Proc. the ACL-02 Workshop on Natural Language Processing in the Biomedical Domain, Philadelphia, USA, July 11, 2002, pp.85-92.
[64] Nguyen N et al. Challenges in pronoun resolution system for biomedical text. In Proc. the Sixth International Language Resources and Evaluation (LREC2008), Marrakech, Morocco, May 28-30, 2008.
[65] Tsai R T H et al. PubMed-EX: A web browser extension to enhance PubMed search with text mining features. Bioinformatics, 2009,
[Epub ahead of print].
[66] Zhang Z et al. Bringing Web 2.0 to bioinformatics. Brief Bioinform., 2009, 10(1): 1-10.
[67] Cheung K et al. Semantic Web Approach to Database Integration in the Life Sciences. Semantic Web: Revolutionizing Knowledge Discovery in the Life Sciences, Springer, 2007, pp.11-30.
[68] Dowell R et al. The distributed annotation system. BMC Bioinformatics, 2001, 2: 7.
[69] O’Reilly T. What is Web 2.0: Design patterns and business models for the next generation of software. 2005, http:// www.oreillynet.com/pub/a/oreilly/tim/news/2005/09/30/ what-is-web-20.html.
[70] Mons B et al. Calling on a million minds for community annotation in WikiProteins. Genome Biology, 2008, 9(5): R89.
[71] Baral C et al. CBioC: Beyond a prototype for collaborative annotation of molecular interactions from the literature. In Proc. Computational Systems Bioinformatics Conference, 2007, 6: 381-384.
[72] Oda K et al. New challenges for text mining: Mapping between text and manually curated pathways. BMC Bioinformatics, 2008, 9(Suppl. 3): S5.
[73] Kanehisa M et al. KEGG for linking genomes to life and the environment. Nucleic Acids Research, 2008, 36(Database Issue): D480-D484.
[74] Hirschman L, Blaschke C. Evaluation of Text Mining in Biology. Text Mining for Biology and Biomedicine, Artech House, 2005, pp.213-245.
[75] Yeh A et al. Background and overview for KDD Cup 2002 task 1: Information extraction from biomedical articles. ACM SIGKDD Explorations Newsletter, 2002, 4(2): 87-89.
[76] HershW, Voorhees E. TREC genomics special issue overview. Information Retrieval, 2009, 12(1): 1-15.
[77] Hakenberg J, Plake C et al. LLL’05 challenge: Genic interaction extraction-identification of language patterns based on alignment and finite state automata. In Proc. the ICML05 Workshop: Learning Language in Logic (LLL05), 2005, 14: 38-45.
[78] Kim J D et al. Overview of BioNLP’09 shared task on event extraction. In Proc. the BioNLP 2009 Workshop Companion Volume for Shared Task, Boulder, USA, June 4-5, 2009, pp.1-9.
[79] Kim J D et al. Corpus annotation for mining biomedical events from literature. BMC Bioinformatics, 2008, 9: 10.
[80] Bader G et al. Pathguide: A pathway resource list. Nucleic Acids Research, 2006, 34(Database Issue): D504-D506.
[81] Camon E et al. The gene ontology annotation (GOA) database: Sharing knowledge in Uniprot with Gene Ontology. Nucleic Acids Research, 2004, 32(Database Issue): D262- D266.
[82] Kim J D et al. GENIA corpus—A semantically annotated corpus for bio-textmining. Bioinformatics, 2003, 19(Suppl. 1): 180-182.
[83] Tanabe L et al. GENETAG: A tagged corpus for gene/protein named entity recognition. BMC Bioinformatics, 2005, 6(Suppl. 1): S3.
[84] Heimonen J et al. Complex-to-pairwise mapping of biological relationships using a semantic network representation. In Proc. the Third International Symposium on Semantic Mining in Biomedicine (SMBM2008), Turku, Finland, Sept. 1-3, 2008, pp.45-52.
[85] Rosario B, Hearst M A. Classifying semantic relations in bioscience texts. In Proc. the 42nd Annual Meeting on Association for Computational Linguistics, Barcelona, Spain, July 21-26, 2004, Article No. 43.
[86] Berleant D et al. Corpus properties of protein interaction descriptions in MEDLINE. 2003, http://class.ee.iastate. edu/berleant/home/me/cv/papers/corpuspropertiesstart. htm.
[87] Nedellec C. Learning language in logic-genic interaction extraction challenge. In Proc. the ICML05 Workshop: Learning Language in Logic (LLL05), Bonn, Germany, Aug. 7, 2005, pp.31-37.
[88] Wattarujeekrit T et al. PASBio: Predicate-argument structures for event extraction in molecular biology. BMC Bioinformatics, Oct. 19, 2004, 5: 155.
[89] Chou W C et al. A semi-automatic method for annotating a biomedical proposition bank. In Proc. ACL Workshop on Frontiers in Linguistically Annotated Corpora, Sydney, Australia, July 22, 2006, pp.5-12.
[90] Seth K et al. Integrated annotation for biomedical information extraction. In Proc. HLT/NAACL-2004, Boston, USA, May 2-7, 2004, pp.61-68.
[91] Tateisi Y, Tsujii J. Part-of-speech annotation of biology research abstracts. In Proc. the 4th International Conference on Language Resource and Evaluation (LREC2004), Lisbon, Portugal, May 26-28, 2004, pp.1267-1270.
[92] Tateisi Y et al. Syntax annotation for the GENIA corpus. In Proc. IJCNLP 2005, Companion Volume, Jeju Island, Korea, Oct. 11-13, 2005, pp.222-227.
[93] Lease M, Charniak E. Parsing biomedical literature. In Proc. the Second International Joint Conference on Natural Language Processing, Jeju Island, Korea, Oct. 11-13, 2005, pp.58- 69.
[94] Smith L et al. MedPost: A part-of-speech tagger for BioMedical text. Bioinformatics, September 22, 2004, 20(14): 2320- 2321.
[95] Krallinger M et al. The BioCreative II.5 challenge overview. In Proc. the BioCreative II.5 Workshop 2009 on Digital Annotations, Madrid, Spain, Oct. 7-9, 2009, p.19.
[96] GasperIn C et al. Annotation of anaphoric relations in biomedical full-text articles using a domain-relevant scheme. In Proc. the Discourse Anaphora and Anaphor Resolution Colloquium, Lagos (Algarve), Portugal, March 29-30, 2007, pp.19-24.
[97] McIntosh M, Curran J. Challenges for automatically extracting molecular interactions from full-text articles. BMC Bioinformatics, 2009, 10: 311.
[98] Kohn K W. Molecular interaction map of the mammalian cell cycle control and DNA repair systems. Mol. Biol. Cell, August 1, 1999, 10(8): 2703-2734.