Master Seminar: Information Retrieval in Web Environments
Software application for prosecution of trademark infringements in the Internet
(working title)
Overview
The European authorities impounded over 250 million units of counterfeit commodities in 2006. The estimated number of unreported cases is multiple times that high. Additionally other types of infringements are common, which can cause major damage to established trademarks. The major part of trademark infringements are conducted via the Internet. The criminal activities usually comprehend:
- Illegal usage of trademark for own, often low-grade products.
- Branded product are stolen and the usual distribution system is annulled.
- Trademark and enterprise image is jeopardized.
The trademark infringements conducted via the Internet hamper their prosecution because of the following features:
- Anonymity: absence of transparency regarding the identity of the supplier.
- Limited Perceptibility of frauds: you find a mixture of phonies, remnants, theft items, second hand articles, trademark infringements and legal sellers.
- Many different forms of communication: professional sales platforms, simple web pages, chat rooms, mailing lists, direct e-mails, etc.
- Delinquency happens all day and night, 24/7.
Therefore a software application is needed which scans various Internet sources and filters them to find indications for possible infringements.
Project Goals
Lawyers and observers usually get the information about possible trademark infringements either by the affected enterprise or by a third party. Much information has to be collected in order to analyse, evaluate and document the facts of the case completely and in a court-proof way. Therefore—in the long run—the intended software application for the prosecution of trademark infringements in the Internet should support the following tasks:
- Crawling of accessible open document sources like product search engines, product repositories, web sides, blogs, news groups etc.
- Crawling of not freely accessible sources like patent dictionaries, commercial registers, chat rooms etc.
- Automated comparison of product offers with existing, official data and product offers (trademarks, price lists, seller lists, etc.) in order to detect possible frauds.
- Storage and processing of results in the face of relevance, validity, etc. (data mining, statistics, reporting, visualisation)
- Matchmaking and similarity analysis of product names and trademarks.
- Compliance with privacy laws is necessary.
Contact
Dr. Dominik Kuropka
dominik.kuropka@hpi.uni-potsdam.de
Room C-2.10
Dates
Topic presentation: 16th of October, 15:15 - 16:45 in Room A-2.2
General meeting: every Tuesday, 15:15 - 16:45 in Room A-2.2
Attachments 
