際際滷

際際滷Share a Scribd company logo

IPO & Ofcom: Measuring IP Rights infringement Research (University of Hertfordshire and AudienceNet)

University of Hertfordshire
University of Hertfordshire

Presentation from September 2013 to Ofcom and the IPO on our research on how to improve measurement of IP Rights infringement including counterfeiting , online copyright, patent and design rights infringements. Further details in the full report published by the IPO https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/325020/IP_Measuring_Infringement.pdf

1 of 14
22/10/2013 1 Measuring Infringement of IPR IPO Report Summary Findings A Review of existing methods and recommendations for new robust methodologies September 2013 Dennis Collopy and Dr Tim Drye Dennis Collopy (UH) and Dr Tim Drye (AN) CONTENT 財 Background 財 Approach Methodology 財 General Observations 財 Distinctive Features 財 Harmonising Framework 財 Application of the Framework Dennis Collopy (UH) and Dr Tim Drye (AN)
22/10/2013 2 IP INFRINGEMENT Dennis Collopy (UH) and Dr Tim Drye (AN) BACKGROUND Study commissioned by the intellectual Property Office (IPO) Scope: Review of methodologies identifying the scale of infringement across 4 main IP Rights Types: Copyright (online & offline) TrademarkPatent Design Rights Dennis Collopy (UH) and Dr Tim Drye (AN)
22/10/2013 3 METHODOLOGY Dennis Collopy (UH) and Dr Tim Drye (AN) METHODOLOGY Team composed of 3 academics from the University of Hertfordshire and 3 Research/Industry experts from Audiencenet 33 48 11 27 38 73 14 22 Online Copyright infringement Patent enforcement Counterfeiting and piracy Design right enforcement reviewed used SOURCES 4 month review of Grey Literature, Trade Body approach to research and views from Experts in the field of piracy, big data, Top down & Bottom Up Approach Classification and segmentation of available research ≒ IP Type ≒ Source of Funding reviewed used reviewed used reviewed used Dennis Collopy (UH) and Dr Tim Drye (AN)
22/10/2013 4 TRADE BODIES RESEARCH Dennis Collopy (UH) and Dr Tim Drye (AN) o Cross sector research compiled from different data sources o Some robust but some not this undermines credibility of the whole (e.g Tera) o ACG advocate approach of using government, industry and consumer data o Industry generated market intelligence (e.g. Fact)- is this collated and made available? o Software, recorded music and film/video bodies carry out some good quality research on online (c) infringement o Commitment to diverse approaches and high standards o Willing to consider alternatives o Even if methodology flawed commitment to consistent and regular measurement (RIAA & BVA/MPAA) o Most trade bodies do not measure scale of infringement - rely on notice and takedown o Evident lack of resources - financial and human o Member surveys - ACID and DACS o One body assembled piracy data from different sources and combined into one statement but based on variable quality data TRADE BODIES RESEARCH Dennis Collopy (UH) and Dr Tim Drye (AN)
22/10/2013 5 TRADE BODIES RESEARCH o Industry research not always about lobbying - also about investment decisions, hence confidential nature of data. o Some research commissioned by trade bodies designed to enable members to react to infringement. o Some trade bodies (PA) provide anti-piracy tools but little evidence the data from this is captured in systemic fashion and available to government. o Resource issues: financial, lack of suitably qualified/ trained staff. o Industry instinct is act quickly fast changing markets greater emphasis on day-to-day anti-piracy measures. o Cannot always wait for government to act. o Dissatisfaction with legal remedies online (e.g. Notice & Takedown) MOTIVES Dennis Collopy (UH) and Dr Tim Drye (AN) EXPERTS VIEWS o MUSICMETRIC Track content using different identification tags.- Data not based on swarms but downloads,- assume all content on Bittorrent 息 infringing. o NIELSEN - Nielsen digital media manager (ndmm) digital watermarking and fingerprinting - aim to provide more reliable way to track content. o BITTORRENT - Data show able to distinguish between different types of content but not track what users do with the torrents, no single server o COUNTY ANALYTICS - BT ran an ad targeting service to monitor user behaviour using DPI. - Main emerging problem in online behaviour namely stream ripping - layers of identification of piracy ; target identification, content verification, container labelling,manual identification, metadata, digital hash, signed metadata, fingerprinting, watermarking - the best piracy identification approaches is layered approach, o BIG CHAMPAGNE - Analysis of torrents. o ONEHOUSE - Problems of measuring ipr infringement online as illustrated by the tor project - defends against a form of network surveillance Not part of the original brief overall sense that future methods for measuring IP infringement more accurately lie as much in technological tools as surveys Dennis Collopy (UH) and Dr Tim Drye (AN)
22/10/2013 6 GENERAL OBSERVATIONS Dennis Collopy (UH) and Dr Tim Drye (AN) The Problem of Divergent Methods and Non-Transparency inhibiting like for like comparisons or methodologies ≒ Lack of methodologies disclosed ≒ Lack of consistency in methodology ≒ Inconsistency between approaches ≒ IPOs Good Evidence Guide mostly absent ≒ Lack of principles & best practice ≒ Lack of coherent Vocabulary and Terminology ≒ No single methodology usable in any of the IPR rights ≒ Survey approach dominant ≒ Nature of Research Objectives: ≒ Ad hoc vs. Longitudinal ≒ Reactive vs. pro-active research ≒ Hypothesis vs. data driven ≒ Commissioning bodies objectives driving research Disconnected approaches within and between segments, -delivery of ad hoc results - general lack of any common overarching reference points both longitudinally or sectorially - little recognition of the perspectives of differing stakeholders. GENERAL OBSERVATIONS Dennis Collopy (UH) and Dr Tim Drye (AN)
22/10/2013 7 DISTINCTIVE FEATURES Dennis Collopy (UH) and Dr Tim Drye (AN) DISTINCTIVE FEATURES Most of the differences arise from the different types of participants, the costs/risks of infringement and recompense: Copyright Trademark Patent Design Victim Content Business Business Business Sector Business Sector Perpetrator Consumer Business Business Sector Business Sector Audience Consumer Consumer Market Market Dennis Collopy (UH) and Dr Tim Drye (AN)
22/10/2013 8 INDUSTRY RESEARCH o Most not of standard suitable for policy making o Most not accompanied by methodology o Some outstanding pieces of research o Consistency & regularity of research crucial o Most research is survey based o Most annual surveys snapshot 'deep dives' not longitudinal o ad hoc surveys commonplace o Question replicability and verifiability o Tech developments like data mining offer possible long-term solutions ONLINE COPYRIGHT ENFORCEMENT DISTINCTIVE FEATURES Dennis Collopy (UH) and Dr Tim Drye (AN) DISTINCTIVE FEATURES 6 main Industry methodologies identifying the scale and value of infringement in different entertainment sectors: o Literature review - collation of secondary data sets e.g. Unifab o Agent-based model e.g. Sandtable o Online surveys e.g. Wiggin and UK Music o Bundled survey approach e.g. Ofcom/Kantar o Mixed approach e.g. data mining (IDC) and survey based (Ipsos) research for BSA o Observational approach - use of various technologies to observe actual behaviour across all forms of online entertainment consumption e.g. NBC/Universal Envisional ONLINE COPYRIGHT ENFORCEMENT Dennis Collopy (UH) and Dr Tim Drye (AN)
22/10/2013 9 GOVERNMENT/ ACADEMIC ≒ None provide comprehensive methodology for input data ≒ Surveys dominate ≒ Focus on consumer behaviour not levels of infringement ≒ Some recognise flaws of survey approach - under-reporting of illicit behaviour ≒ Recommend observational approach ≒ Most academic studies cannot be generalised few using representative samples ONLINE COPYRIGHT ENFORCEMENT DISTINCTIVE FEATURES Dennis Collopy (UH) and Dr Tim Drye (AN) INDUSTRY o Harder still to measure infringement levels o Reliance on consumer surveys o Literature focus on broader issues of C&P-result of commissioning process o Adoption of CEBR 2000 'omnibus approach' o Standard measurements post trips based on customs seizures o Mark Monitor approach - has potential but little recognition of skewed nature of sample-best for individual brand GOVERNMENT & ACADEMIC oHeavy reliance on customs seizures - mere fraction of total oOECD correlated industry and seizure data but recognise this is crude oOECD suggest combining objective and robust methodologies oBut little systematic collection and evaluation of data oSome rely on anecdotal and fragmentary info oEvident a mixed approach on hard data from government, industry and consumers oOHIM observatory note the 'snapshot' nature of most studies oRand approach includes four solid approaches but 5th approach an economic model based on unreliable data set. COUNTERFEITING & PIRACY (C & P) DISTINCTIVE FEATURES Dennis Collopy (UH) and Dr Tim Drye (AN)
22/10/2013 10 INDUSTRY 財 Discernible lack of industry generated research 財 Competitive strategies in industries dominated by small number big players in tech and pharma 財 Porters five forces apply 財 Microsoft IP license strategy post 2003 - aim to stop litigation and build relationships GOVERNMENT 財 Bentley et al (2009) mixed methodology survey firms and sampling court cases 財 Benefits of cooperation with industry and IP practioners to improve verification 財 But this still only captures fraction of market PATENT ENFORCEMENT DISTINCTIVE FEATURES Dennis Collopy (UH) and Dr Tim Drye (AN) DISTINCTIVE FEATURES ACADEMIC 財Real insights based on analysis of litigation nb UK and USA 財Litigated patents greater value than non-litigated patents (Helmers & McDonagh) 財NPEs responsible for 40% (from 22% over 5 years) of patent infringement lawsuits filed in the us( Jeruss & Feldman et al) 財Excluding pharma & chemistry industries, costs incurred litigating patents outweigh the earnings gained 財Internet-related patents litigated 7.5 to 9.5 times more frequently than non-internet patents in the software industry. 財High litigation costs benefit larger firms(Schliessler) 財Increasing importance of compulsory licensing over injunctive relief in US decision post eBay v mercexchange (Venkatesan) 財Significant amount of infringement activity hidden from measurement (settlements / cross licensing) 財Weatherall & Webster 2010 inventors survey builds on the 2009 model proposed by Bentley , Weatherall & Webster 財Model viable within other IPR (see designs) PATENT ENFORCEMENT Dennis Collopy (UH) and Dr Tim Drye (AN)
22/10/2013 11 DESIGN RIGHT ENFORCEMENT 財 Little relevant literature - some government & academic 財 Complexity (patchwork) of rights for designers includes some (c) remedies 財 Existing litigation costs inhibit infringement actions by individual designers and sme's 財 Possible changes if Carter-Silk and Lewistons 2012 recommendations implemented 財 ACID survey not fully representative but good basis for methodology to measure infringement 財 Possible use of customs data but same caveats apply about representation of infringing activity 財 Could integrate with Weatherall & Webster patent model to survey designers 財 But must use in conjunction with other public data sources DISTINCTIVE FEATURES Dennis Collopy (UH) and Dr Tim Drye (AN) CONCEPTUAL FRAMEWORK Dennis Collopy (UH) and Dr Tim Drye (AN)
22/10/2013 12 1. Highly skewed behaviour 2. Dynamically changing perceptions 3. Rapidly changing costs of distribution and recompense 4. Driven by perception of risk CONCEPTUAL FRAMEWORK Dennis Collopy (UH) and Dr Tim Drye (AN) APPLICATION OF THE FRAMEWORK Dennis Collopy (UH) and Dr Tim Drye (AN)
22/10/2013 13 APPLICATION OF THE FRAMEWORK PUBLIC REFERENCE POINT AUDIENCE CREATOR PERPETRATOR Dennis Collopy (UH) and Dr Tim Drye (AN) THE CONCLUSION: A CALL FOR CONSISTENT OVERSIGHT Dennis Collopy (UH) and Dr Tim Drye (AN)
22/10/2013 14 A call for the need to establish a consensus around the common stack in the ground, a role in this instance for a government contribution to act as the guardian and frame of reference, this will require a commitment to a sustained and regular reporting of a sampling of the participants, alongside longitudinal collation of extreme behaviour. CONCLUSION Dennis Collopy (UH) and Dr Tim Drye (AN)

More Related Content

IPO & Ofcom: Measuring IP Rights infringement Research (University of Hertfordshire and AudienceNet)

  • 1. 22/10/2013 1 Measuring Infringement of IPR IPO Report Summary Findings A Review of existing methods and recommendations for new robust methodologies September 2013 Dennis Collopy and Dr Tim Drye Dennis Collopy (UH) and Dr Tim Drye (AN) CONTENT 財 Background 財 Approach Methodology 財 General Observations 財 Distinctive Features 財 Harmonising Framework 財 Application of the Framework Dennis Collopy (UH) and Dr Tim Drye (AN)
  • 2. 22/10/2013 2 IP INFRINGEMENT Dennis Collopy (UH) and Dr Tim Drye (AN) BACKGROUND Study commissioned by the intellectual Property Office (IPO) Scope: Review of methodologies identifying the scale of infringement across 4 main IP Rights Types: Copyright (online & offline) TrademarkPatent Design Rights Dennis Collopy (UH) and Dr Tim Drye (AN)
  • 3. 22/10/2013 3 METHODOLOGY Dennis Collopy (UH) and Dr Tim Drye (AN) METHODOLOGY Team composed of 3 academics from the University of Hertfordshire and 3 Research/Industry experts from Audiencenet 33 48 11 27 38 73 14 22 Online Copyright infringement Patent enforcement Counterfeiting and piracy Design right enforcement reviewed used SOURCES 4 month review of Grey Literature, Trade Body approach to research and views from Experts in the field of piracy, big data, Top down & Bottom Up Approach Classification and segmentation of available research ≒ IP Type ≒ Source of Funding reviewed used reviewed used reviewed used Dennis Collopy (UH) and Dr Tim Drye (AN)
  • 4. 22/10/2013 4 TRADE BODIES RESEARCH Dennis Collopy (UH) and Dr Tim Drye (AN) o Cross sector research compiled from different data sources o Some robust but some not this undermines credibility of the whole (e.g Tera) o ACG advocate approach of using government, industry and consumer data o Industry generated market intelligence (e.g. Fact)- is this collated and made available? o Software, recorded music and film/video bodies carry out some good quality research on online (c) infringement o Commitment to diverse approaches and high standards o Willing to consider alternatives o Even if methodology flawed commitment to consistent and regular measurement (RIAA & BVA/MPAA) o Most trade bodies do not measure scale of infringement - rely on notice and takedown o Evident lack of resources - financial and human o Member surveys - ACID and DACS o One body assembled piracy data from different sources and combined into one statement but based on variable quality data TRADE BODIES RESEARCH Dennis Collopy (UH) and Dr Tim Drye (AN)
  • 5. 22/10/2013 5 TRADE BODIES RESEARCH o Industry research not always about lobbying - also about investment decisions, hence confidential nature of data. o Some research commissioned by trade bodies designed to enable members to react to infringement. o Some trade bodies (PA) provide anti-piracy tools but little evidence the data from this is captured in systemic fashion and available to government. o Resource issues: financial, lack of suitably qualified/ trained staff. o Industry instinct is act quickly fast changing markets greater emphasis on day-to-day anti-piracy measures. o Cannot always wait for government to act. o Dissatisfaction with legal remedies online (e.g. Notice & Takedown) MOTIVES Dennis Collopy (UH) and Dr Tim Drye (AN) EXPERTS VIEWS o MUSICMETRIC Track content using different identification tags.- Data not based on swarms but downloads,- assume all content on Bittorrent 息 infringing. o NIELSEN - Nielsen digital media manager (ndmm) digital watermarking and fingerprinting - aim to provide more reliable way to track content. o BITTORRENT - Data show able to distinguish between different types of content but not track what users do with the torrents, no single server o COUNTY ANALYTICS - BT ran an ad targeting service to monitor user behaviour using DPI. - Main emerging problem in online behaviour namely stream ripping - layers of identification of piracy ; target identification, content verification, container labelling,manual identification, metadata, digital hash, signed metadata, fingerprinting, watermarking - the best piracy identification approaches is layered approach, o BIG CHAMPAGNE - Analysis of torrents. o ONEHOUSE - Problems of measuring ipr infringement online as illustrated by the tor project - defends against a form of network surveillance Not part of the original brief overall sense that future methods for measuring IP infringement more accurately lie as much in technological tools as surveys Dennis Collopy (UH) and Dr Tim Drye (AN)
  • 6. 22/10/2013 6 GENERAL OBSERVATIONS Dennis Collopy (UH) and Dr Tim Drye (AN) The Problem of Divergent Methods and Non-Transparency inhibiting like for like comparisons or methodologies ≒ Lack of methodologies disclosed ≒ Lack of consistency in methodology ≒ Inconsistency between approaches ≒ IPOs Good Evidence Guide mostly absent ≒ Lack of principles & best practice ≒ Lack of coherent Vocabulary and Terminology ≒ No single methodology usable in any of the IPR rights ≒ Survey approach dominant ≒ Nature of Research Objectives: ≒ Ad hoc vs. Longitudinal ≒ Reactive vs. pro-active research ≒ Hypothesis vs. data driven ≒ Commissioning bodies objectives driving research Disconnected approaches within and between segments, -delivery of ad hoc results - general lack of any common overarching reference points both longitudinally or sectorially - little recognition of the perspectives of differing stakeholders. GENERAL OBSERVATIONS Dennis Collopy (UH) and Dr Tim Drye (AN)
  • 7. 22/10/2013 7 DISTINCTIVE FEATURES Dennis Collopy (UH) and Dr Tim Drye (AN) DISTINCTIVE FEATURES Most of the differences arise from the different types of participants, the costs/risks of infringement and recompense: Copyright Trademark Patent Design Victim Content Business Business Business Sector Business Sector Perpetrator Consumer Business Business Sector Business Sector Audience Consumer Consumer Market Market Dennis Collopy (UH) and Dr Tim Drye (AN)
  • 8. 22/10/2013 8 INDUSTRY RESEARCH o Most not of standard suitable for policy making o Most not accompanied by methodology o Some outstanding pieces of research o Consistency & regularity of research crucial o Most research is survey based o Most annual surveys snapshot 'deep dives' not longitudinal o ad hoc surveys commonplace o Question replicability and verifiability o Tech developments like data mining offer possible long-term solutions ONLINE COPYRIGHT ENFORCEMENT DISTINCTIVE FEATURES Dennis Collopy (UH) and Dr Tim Drye (AN) DISTINCTIVE FEATURES 6 main Industry methodologies identifying the scale and value of infringement in different entertainment sectors: o Literature review - collation of secondary data sets e.g. Unifab o Agent-based model e.g. Sandtable o Online surveys e.g. Wiggin and UK Music o Bundled survey approach e.g. Ofcom/Kantar o Mixed approach e.g. data mining (IDC) and survey based (Ipsos) research for BSA o Observational approach - use of various technologies to observe actual behaviour across all forms of online entertainment consumption e.g. NBC/Universal Envisional ONLINE COPYRIGHT ENFORCEMENT Dennis Collopy (UH) and Dr Tim Drye (AN)
  • 9. 22/10/2013 9 GOVERNMENT/ ACADEMIC ≒ None provide comprehensive methodology for input data ≒ Surveys dominate ≒ Focus on consumer behaviour not levels of infringement ≒ Some recognise flaws of survey approach - under-reporting of illicit behaviour ≒ Recommend observational approach ≒ Most academic studies cannot be generalised few using representative samples ONLINE COPYRIGHT ENFORCEMENT DISTINCTIVE FEATURES Dennis Collopy (UH) and Dr Tim Drye (AN) INDUSTRY o Harder still to measure infringement levels o Reliance on consumer surveys o Literature focus on broader issues of C&P-result of commissioning process o Adoption of CEBR 2000 'omnibus approach' o Standard measurements post trips based on customs seizures o Mark Monitor approach - has potential but little recognition of skewed nature of sample-best for individual brand GOVERNMENT & ACADEMIC oHeavy reliance on customs seizures - mere fraction of total oOECD correlated industry and seizure data but recognise this is crude oOECD suggest combining objective and robust methodologies oBut little systematic collection and evaluation of data oSome rely on anecdotal and fragmentary info oEvident a mixed approach on hard data from government, industry and consumers oOHIM observatory note the 'snapshot' nature of most studies oRand approach includes four solid approaches but 5th approach an economic model based on unreliable data set. COUNTERFEITING & PIRACY (C & P) DISTINCTIVE FEATURES Dennis Collopy (UH) and Dr Tim Drye (AN)
  • 10. 22/10/2013 10 INDUSTRY 財 Discernible lack of industry generated research 財 Competitive strategies in industries dominated by small number big players in tech and pharma 財 Porters five forces apply 財 Microsoft IP license strategy post 2003 - aim to stop litigation and build relationships GOVERNMENT 財 Bentley et al (2009) mixed methodology survey firms and sampling court cases 財 Benefits of cooperation with industry and IP practioners to improve verification 財 But this still only captures fraction of market PATENT ENFORCEMENT DISTINCTIVE FEATURES Dennis Collopy (UH) and Dr Tim Drye (AN) DISTINCTIVE FEATURES ACADEMIC 財Real insights based on analysis of litigation nb UK and USA 財Litigated patents greater value than non-litigated patents (Helmers & McDonagh) 財NPEs responsible for 40% (from 22% over 5 years) of patent infringement lawsuits filed in the us( Jeruss & Feldman et al) 財Excluding pharma & chemistry industries, costs incurred litigating patents outweigh the earnings gained 財Internet-related patents litigated 7.5 to 9.5 times more frequently than non-internet patents in the software industry. 財High litigation costs benefit larger firms(Schliessler) 財Increasing importance of compulsory licensing over injunctive relief in US decision post eBay v mercexchange (Venkatesan) 財Significant amount of infringement activity hidden from measurement (settlements / cross licensing) 財Weatherall & Webster 2010 inventors survey builds on the 2009 model proposed by Bentley , Weatherall & Webster 財Model viable within other IPR (see designs) PATENT ENFORCEMENT Dennis Collopy (UH) and Dr Tim Drye (AN)
  • 11. 22/10/2013 11 DESIGN RIGHT ENFORCEMENT 財 Little relevant literature - some government & academic 財 Complexity (patchwork) of rights for designers includes some (c) remedies 財 Existing litigation costs inhibit infringement actions by individual designers and sme's 財 Possible changes if Carter-Silk and Lewistons 2012 recommendations implemented 財 ACID survey not fully representative but good basis for methodology to measure infringement 財 Possible use of customs data but same caveats apply about representation of infringing activity 財 Could integrate with Weatherall & Webster patent model to survey designers 財 But must use in conjunction with other public data sources DISTINCTIVE FEATURES Dennis Collopy (UH) and Dr Tim Drye (AN) CONCEPTUAL FRAMEWORK Dennis Collopy (UH) and Dr Tim Drye (AN)
  • 12. 22/10/2013 12 1. Highly skewed behaviour 2. Dynamically changing perceptions 3. Rapidly changing costs of distribution and recompense 4. Driven by perception of risk CONCEPTUAL FRAMEWORK Dennis Collopy (UH) and Dr Tim Drye (AN) APPLICATION OF THE FRAMEWORK Dennis Collopy (UH) and Dr Tim Drye (AN)
  • 13. 22/10/2013 13 APPLICATION OF THE FRAMEWORK PUBLIC REFERENCE POINT AUDIENCE CREATOR PERPETRATOR Dennis Collopy (UH) and Dr Tim Drye (AN) THE CONCLUSION: A CALL FOR CONSISTENT OVERSIGHT Dennis Collopy (UH) and Dr Tim Drye (AN)
  • 14. 22/10/2013 14 A call for the need to establish a consensus around the common stack in the ground, a role in this instance for a government contribution to act as the guardian and frame of reference, this will require a commitment to a sustained and regular reporting of a sampling of the participants, alongside longitudinal collation of extreme behaviour. CONCLUSION Dennis Collopy (UH) and Dr Tim Drye (AN)