Exploiting evidence from unstructured data to enhance master data management

    Research output: Contribution to journalArticle

    Published

    Standard

    Exploiting evidence from unstructured data to enhance master data management. / Murthy, Karin; Deshpande, Prasad; Dey, Atreyee; Halasipuram, Ramanujam; Mohania, Mukesh; Padmanabhan, Deepak; Reed, Jennifer; Schumacher, Scott.

    In: Proceedings of the VLDB Endowment, Vol. 5, No. 12, 2012, p. 1862-1873.

    Research output: Contribution to journalArticle

    Harvard

    Murthy, K, Deshpande, P, Dey, A, Halasipuram, R, Mohania, M, Padmanabhan, D, Reed, J & Schumacher, S 2012, 'Exploiting evidence from unstructured data to enhance master data management', Proceedings of the VLDB Endowment, vol. 5, no. 12, pp. 1862-1873.

    APA

    Murthy, K., Deshpande, P., Dey, A., Halasipuram, R., Mohania, M., Padmanabhan, D., ... Schumacher, S. (2012). Exploiting evidence from unstructured data to enhance master data management. Proceedings of the VLDB Endowment, 5(12), 1862-1873.

    Vancouver

    Murthy K, Deshpande P, Dey A, Halasipuram R, Mohania M, Padmanabhan D et al. Exploiting evidence from unstructured data to enhance master data management. Proceedings of the VLDB Endowment. 2012;5(12):1862-1873.

    Author

    Murthy, Karin ; Deshpande, Prasad ; Dey, Atreyee ; Halasipuram, Ramanujam ; Mohania, Mukesh ; Padmanabhan, Deepak ; Reed, Jennifer ; Schumacher, Scott. / Exploiting evidence from unstructured data to enhance master data management. In: Proceedings of the VLDB Endowment. 2012 ; Vol. 5, No. 12. pp. 1862-1873.

    Bibtex

    @article{601e24daa5804efeaabae75940a84aa8,
    title = "Exploiting evidence from unstructured data to enhance master data management",
    abstract = "Master data management (MDM) integrates data from multiplestructured data sources and builds a consolidated 360-degree view of business entities such as customers and products.Today’s MDM systems are not prepared to integrateinformation from unstructured data sources, such as newsreports, emails, call-center transcripts, and chat logs. However,those unstructured data sources may contain valuableinformation about the same entities known to MDM fromthe structured data sources. Integrating information fromunstructured data into MDM is challenging as textual referencesto existing MDM entities are often incomplete andimprecise and the additional entity information extractedfrom text should not impact the trustworthiness of MDMdata.In this paper, we present an architecture for making MDMtext-aware and showcase its implementation as IBM InfoSphereMDM Extension for Unstructured Text Correlation,an add-on to IBM InfoSphere Master Data ManagementStandard Edition. We highlight how MDM benefits fromadditional evidence found in documents when doing entityresolution and relationship discovery. We experimentallydemonstrate the feasibility of integrating information fromunstructured data sources into MDM.",
    author = "Karin Murthy and Prasad Deshpande and Atreyee Dey and Ramanujam Halasipuram and Mukesh Mohania and Deepak Padmanabhan and Jennifer Reed and Scott Schumacher",
    year = "2012",
    language = "English",
    volume = "5",
    pages = "1862--1873",
    journal = "Proceedings of the VLDB Endowment",
    issn = "2150-8097",
    publisher = "Very Large Data Base Endowment Inc.",
    number = "12",

    }

    RIS

    TY - JOUR

    T1 - Exploiting evidence from unstructured data to enhance master data management

    AU - Murthy, Karin

    AU - Deshpande, Prasad

    AU - Dey, Atreyee

    AU - Halasipuram, Ramanujam

    AU - Mohania, Mukesh

    AU - Padmanabhan, Deepak

    AU - Reed, Jennifer

    AU - Schumacher, Scott

    PY - 2012

    Y1 - 2012

    N2 - Master data management (MDM) integrates data from multiplestructured data sources and builds a consolidated 360-degree view of business entities such as customers and products.Today’s MDM systems are not prepared to integrateinformation from unstructured data sources, such as newsreports, emails, call-center transcripts, and chat logs. However,those unstructured data sources may contain valuableinformation about the same entities known to MDM fromthe structured data sources. Integrating information fromunstructured data into MDM is challenging as textual referencesto existing MDM entities are often incomplete andimprecise and the additional entity information extractedfrom text should not impact the trustworthiness of MDMdata.In this paper, we present an architecture for making MDMtext-aware and showcase its implementation as IBM InfoSphereMDM Extension for Unstructured Text Correlation,an add-on to IBM InfoSphere Master Data ManagementStandard Edition. We highlight how MDM benefits fromadditional evidence found in documents when doing entityresolution and relationship discovery. We experimentallydemonstrate the feasibility of integrating information fromunstructured data sources into MDM.

    AB - Master data management (MDM) integrates data from multiplestructured data sources and builds a consolidated 360-degree view of business entities such as customers and products.Today’s MDM systems are not prepared to integrateinformation from unstructured data sources, such as newsreports, emails, call-center transcripts, and chat logs. However,those unstructured data sources may contain valuableinformation about the same entities known to MDM fromthe structured data sources. Integrating information fromunstructured data into MDM is challenging as textual referencesto existing MDM entities are often incomplete andimprecise and the additional entity information extractedfrom text should not impact the trustworthiness of MDMdata.In this paper, we present an architecture for making MDMtext-aware and showcase its implementation as IBM InfoSphereMDM Extension for Unstructured Text Correlation,an add-on to IBM InfoSphere Master Data ManagementStandard Edition. We highlight how MDM benefits fromadditional evidence found in documents when doing entityresolution and relationship discovery. We experimentallydemonstrate the feasibility of integrating information fromunstructured data sources into MDM.

    M3 - Article

    VL - 5

    SP - 1862

    EP - 1873

    JO - Proceedings of the VLDB Endowment

    T2 - Proceedings of the VLDB Endowment

    JF - Proceedings of the VLDB Endowment

    SN - 2150-8097

    IS - 12

    ER -

    Download

    Download as: RIS