{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"code","source":"# This Python 3 environment comes with many helpful analytics libraries installed\n# It is defined by the kaggle/python Docker image: https://github.com/kaggle/docker-python\n# For example, here's several helpful packages to load\n\nimport numpy as np # linear algebra\nimport pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\n\n# Input data files are available in the read-only \"../input/\" directory\n# For example, running this (by clicking run or pressing Shift+Enter) will list all files under the input directory\n\nimport os\nfor dirname, _, filenames in os.walk('/kaggle/input'):\n    for filename in filenames:\n        print(os.path.join(dirname, filename))\n\n# You can write up to 20GB to the current directory (/kaggle/working/) that gets preserved as output when you create a version using \"Save & Run All\" \n# You can also write temporary files to /kaggle/temp/, but they won't be saved outside of the current session","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","execution":{"iopub.status.busy":"2023-07-02T21:55:55.616496Z","iopub.execute_input":"2023-07-02T21:55:55.616925Z","iopub.status.idle":"2023-07-02T21:55:55.625468Z","shell.execute_reply.started":"2023-07-02T21:55:55.616863Z","shell.execute_reply":"2023-07-02T21:55:55.624208Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"Using the U.S. Patent Office Databases to Track AI Developments\n","metadata":{}},{"cell_type":"markdown","source":"Introduction\n\n   The US Patent Office is a library of technology ideas, including artificial intelligence (AI).  The U.S. Patent Office databases are organized and are constantly updated and are therefore a good means of tracking developments in AI.  Both the benefits and the flaw using the patent system for this purpose will be discussed.","metadata":{}},{"cell_type":"markdown","source":"The US Patent System\n\n    The legal basis for the U.S. patent system is found in the Constitution so the idea of such a repository was part of the founding of the country.  The Patent Office was established in the early 1800s, making it one of the oldest institutions in the U.S.  The more often spoken-about aspect of the patent system is the aspect of granting a patent owner a period of monopoly for making the invention and then making piles of money from selling it or from suing others selling the invention for infringement.  The less often spoken-about aspect is that the patent system as the repository of technology ideas which inventors have disclosed for recording in the repository in exchange for the monopoly.  The US Patent Office organizes these ideas into classes and subclasses of technology which enables particular technology ideas to be more easily and accurately found.","metadata":{}},{"cell_type":"markdown","source":"AI is patentable and U.S. Patent Office AI Diffusion Study\n\n       Artificial intelligence technology has been eligible for patent for many years although the patents containing AI are not always easy to identify.  Unlike some fields, such as chemistry, software terminology, and by extension AI terminology, is not universally defined and agree-upon which can make word-based searching for AI difficult.  Also the older laws and rules for patent applications discouraged the use of certain terminology that would make the presence of AI in a patent more apparent.  In addition, a patent applicant is allowed to be his or her own lexicographer which means and inventor can make up words or expand the definition of existing words to describe the technology in a patent application which further complicates the terminology in software technologies and AI.  Happily, the laws and rules for AI patents have changed for the better, although reasonable minds may differ on how much the situation has improved.\t\n\n        The US Patent Office studied patents between the years 1976 and 2018 for the presence of AI.  The study included the volume, nature and evolution of AI in patents.  The results were published in a paper in October 2020.  Machine learning was used to uncover the presence of AI in the patents.  Notably the study found that annual AI patent applications to the Patent Office increased by more than 100% between the years 2002 to 2018.  In 2018, more than 60,000 AI patent applications were filed.  In the same year, more than 16% of all patent applications had some AI elements.  This data should obviate any doubt that the US Patent Office databases contain insufficient material for the study of the state of AI technologies.\n\t\t\t\t\t\t","metadata":{}},{"cell_type":"markdown","source":"Advantages\t\n\t\t\t\t\t\n     There are a number of advantages in using U.S. patents and published patent applications over published academic papers for study of the state of AI.  Disclosure requirements for inventions generally make the technology more readily accessible than in academic papers.\t\n\t\t\n     The problem -solution format of patents enables someone seeking to solve a particular problem find at least one solution, if it exists in the repository, and if a number of groups are working on solutions, probably many solutions.\n\n      The patent databases are organized systems where technologies are filed into classes and subclasses.  Patent information includes dated submissions and dated publications and dates of patent issue, along with lists of inventors and patent owners.  As stated above, terminology in software and by extension AI is inconsistent making it difficult to word search – classification makes it easier.\n\n      One more advantage of a patent review to review of academic papers is that a patent typically indicates a strong possibility of commercialization and therefore a technology that is likely to have some impact on the public.  Patents also make it possible to uncover  commercial trends.","metadata":{}},{"cell_type":"markdown","source":"Limitations\n\n            There are limitations to using U.S. patents and published patent applications to study the state of AI.\n\n\tThere can be a considerable time lag between the date of filing a patent application and the date a patent issues and becomes available for public view.  Patent applications are published after eighteen months, however, not all patent applications are published.\n\n\tInventors and invention owners are not required to file for patent.  Accordingly, only inventors or owners who believe in the patent system will file for patent and therefore be present in the patent databases.  Further, patent applications tend to be very expensive particularly in AI and so only those who have sufficient funds to support a patent application are able to file for patent.\n\n\tThis paper is limited to filings in the US patent office but this does not mean that the technologies are only U.S.-based as foreign entities are allowed to apply for patent and they do apply for patent in the U.S.\n","metadata":{}},{"cell_type":"markdown","source":"Alternatives to the U.S. Patent Office\n\n          U.S. patents are available for searching in other databases.  The other databases are typically commercial in one way or another and while some have an up-front cost typically in the form of a subscription cost they all involve taking in and possibly using search query data in ways that a researcher may not want.  Use of the U.S. patent databases is free and the search query data is relatively secure.\n\n          The databases of the World Intellectual Property Organization (WIPO) include U.S. patents and are also free to use.  The WIPO databases also include patent applications from many countries besides the U.S.","metadata":{}},{"cell_type":"markdown","source":"Conclusion\n\n         The data and tools for study of the state of AI are available in the U.S. Patent Office.  The volume and variety of technologies available for study provides a reasonably reliable view of the state of AI in many areas.  The organization of the patent ideas repository provides some advantages over a study of academic papers for some state-of-the art research.","metadata":{}},{"cell_type":"markdown","source":"References\n\n1.  Inventing AI: Tracing the Diffusion of Artificial Intelligence with U.S. Patents, U.S. Patent and Trademark Office • Office of the Chief Economist IP DATA HIGHLIGHTS • Number 5, October 2020 \t\t\n\n2.  United States Patent and Trademark Office, www.uspto.gov.\n\n3.  World Intellectual Property Organization, www.wipo.org. ","metadata":{}},{"cell_type":"code","source":"","metadata":{"execution":{"iopub.status.busy":"2023-07-02T22:10:10.283688Z","iopub.execute_input":"2023-07-02T22:10:10.284117Z","iopub.status.idle":"2023-07-02T22:10:10.6676Z","shell.execute_reply.started":"2023-07-02T22:10:10.284086Z","shell.execute_reply":"2023-07-02T22:10:10.666278Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"","metadata":{"trusted":true},"execution_count":null,"outputs":[]}]}