Firms' knowledge profiles: Mapping patent data with unsupervised learning

Arho Suominen (Corresponding Author), Hannes Toivanen, Marko Seppänen

    Research output: Contribution to journalArticleScientificpeer-review

    18 Citations (Scopus)

    Abstract

    Patent data has been an obvious choice for analysis leading to strategic technology intelligence, yet, the recent proliferation of machine learning text analysis methods is changing the status of traditional patent data analysis methods and approaches. This article discusses the benefits and constraints of machine learning approaches in industry level patent analysis, and to this end offers a demonstration of unsupervised learning based analysis of the leading telecommunication firms between 2001 and 2014 based on about 160,000 USPTO full-text patents. Data were classified using full-text descriptions with Latent Dirichlet Allocation, and latent patterns emerging through the unsupervised learning process were modelled by company and year to create an overall view of patenting within the industry, and to forecast future trends. Our results demonstrate company-specific differences in their knowledge profiles, as well as show the evolution of the knowledge profiles of industry leaders from hardware to software focussed technology strategies. The results cast also light on the dynamics of emerging and declining knowledge areas in the telecommunication industry. Our results prompt a consideration of the current status of established approaches to patent landscaping, such as key-word or technology classifications and other approaches relying on semantic labelling, in the context of novel machine learning approaches. Finally, we discuss implications for policy makers, and, in particular, for strategic management in firms.
    Original languageEnglish
    Pages (from-to)131-142
    Number of pages12
    JournalTechnological Forecasting and Social Change
    Volume115
    DOIs
    Publication statusPublished - 1 Feb 2017
    MoE publication typeA1 Journal article-refereed

    Fingerprint

    Unsupervised learning
    Industry
    Learning
    Telecommunications
    Learning systems
    Technology
    Patents
    Telecommunication industry
    Administrative Personnel
    Intelligence
    Semantics
    Software
    Labeling
    Telecommunication
    Demonstrations
    Patent data
    Hardware
    Machine Learning
    Machine learning

    Keywords

    • technology management
    • patent analysis
    • unsupervised learning
    • topic modelling
    • telecommunication industry

    Cite this

    @article{96d7170fa8be44c3ad8932263da8171a,
    title = "Firms' knowledge profiles: Mapping patent data with unsupervised learning",
    abstract = "Patent data has been an obvious choice for analysis leading to strategic technology intelligence, yet, the recent proliferation of machine learning text analysis methods is changing the status of traditional patent data analysis methods and approaches. This article discusses the benefits and constraints of machine learning approaches in industry level patent analysis, and to this end offers a demonstration of unsupervised learning based analysis of the leading telecommunication firms between 2001 and 2014 based on about 160,000 USPTO full-text patents. Data were classified using full-text descriptions with Latent Dirichlet Allocation, and latent patterns emerging through the unsupervised learning process were modelled by company and year to create an overall view of patenting within the industry, and to forecast future trends. Our results demonstrate company-specific differences in their knowledge profiles, as well as show the evolution of the knowledge profiles of industry leaders from hardware to software focussed technology strategies. The results cast also light on the dynamics of emerging and declining knowledge areas in the telecommunication industry. Our results prompt a consideration of the current status of established approaches to patent landscaping, such as key-word or technology classifications and other approaches relying on semantic labelling, in the context of novel machine learning approaches. Finally, we discuss implications for policy makers, and, in particular, for strategic management in firms.",
    keywords = "technology management, patent analysis, unsupervised learning, topic modelling, telecommunication industry",
    author = "Arho Suominen and Hannes Toivanen and Marko Sepp{\"a}nen",
    note = "Project code: 102296 ISI: BUSINESS",
    year = "2017",
    month = "2",
    day = "1",
    doi = "10.1016/j.techfore.2016.09.028",
    language = "English",
    volume = "115",
    pages = "131--142",
    journal = "Technological Forecasting and Social Change",
    issn = "0040-1625",
    publisher = "Elsevier",

    }

    Firms' knowledge profiles : Mapping patent data with unsupervised learning. / Suominen, Arho (Corresponding Author); Toivanen, Hannes; Seppänen, Marko.

    In: Technological Forecasting and Social Change, Vol. 115, 01.02.2017, p. 131-142.

    Research output: Contribution to journalArticleScientificpeer-review

    TY - JOUR

    T1 - Firms' knowledge profiles

    T2 - Mapping patent data with unsupervised learning

    AU - Suominen, Arho

    AU - Toivanen, Hannes

    AU - Seppänen, Marko

    N1 - Project code: 102296 ISI: BUSINESS

    PY - 2017/2/1

    Y1 - 2017/2/1

    N2 - Patent data has been an obvious choice for analysis leading to strategic technology intelligence, yet, the recent proliferation of machine learning text analysis methods is changing the status of traditional patent data analysis methods and approaches. This article discusses the benefits and constraints of machine learning approaches in industry level patent analysis, and to this end offers a demonstration of unsupervised learning based analysis of the leading telecommunication firms between 2001 and 2014 based on about 160,000 USPTO full-text patents. Data were classified using full-text descriptions with Latent Dirichlet Allocation, and latent patterns emerging through the unsupervised learning process were modelled by company and year to create an overall view of patenting within the industry, and to forecast future trends. Our results demonstrate company-specific differences in their knowledge profiles, as well as show the evolution of the knowledge profiles of industry leaders from hardware to software focussed technology strategies. The results cast also light on the dynamics of emerging and declining knowledge areas in the telecommunication industry. Our results prompt a consideration of the current status of established approaches to patent landscaping, such as key-word or technology classifications and other approaches relying on semantic labelling, in the context of novel machine learning approaches. Finally, we discuss implications for policy makers, and, in particular, for strategic management in firms.

    AB - Patent data has been an obvious choice for analysis leading to strategic technology intelligence, yet, the recent proliferation of machine learning text analysis methods is changing the status of traditional patent data analysis methods and approaches. This article discusses the benefits and constraints of machine learning approaches in industry level patent analysis, and to this end offers a demonstration of unsupervised learning based analysis of the leading telecommunication firms between 2001 and 2014 based on about 160,000 USPTO full-text patents. Data were classified using full-text descriptions with Latent Dirichlet Allocation, and latent patterns emerging through the unsupervised learning process were modelled by company and year to create an overall view of patenting within the industry, and to forecast future trends. Our results demonstrate company-specific differences in their knowledge profiles, as well as show the evolution of the knowledge profiles of industry leaders from hardware to software focussed technology strategies. The results cast also light on the dynamics of emerging and declining knowledge areas in the telecommunication industry. Our results prompt a consideration of the current status of established approaches to patent landscaping, such as key-word or technology classifications and other approaches relying on semantic labelling, in the context of novel machine learning approaches. Finally, we discuss implications for policy makers, and, in particular, for strategic management in firms.

    KW - technology management

    KW - patent analysis

    KW - unsupervised learning

    KW - topic modelling

    KW - telecommunication industry

    UR - http://www.scopus.com/inward/record.url?scp=84992118302&partnerID=8YFLogxK

    U2 - 10.1016/j.techfore.2016.09.028

    DO - 10.1016/j.techfore.2016.09.028

    M3 - Article

    VL - 115

    SP - 131

    EP - 142

    JO - Technological Forecasting and Social Change

    JF - Technological Forecasting and Social Change

    SN - 0040-1625

    ER -