Tuesday 6. October 2020

The report, which takes a focus on Artificial Intelligence (AI), looks at the expertise held by Canadian researchers and institutions undertaken both domestically and abroad in this continuously evolving technology field.  Artificial Intelligence (AI) is a technology area that has garnered significant interest in recent years, however measuring innovation pertaining to AI is a challenging task since the field involves a variety of different techniques that can be broadly applied across a wide array of industries.  The findings reveal Canada ranks sixth globally, both in terms of the number of patented inventions assigned to Canadian researchers and to Canadian institutions. Canada’s rankings fall behind notable countries that file prolifically, namely China and the United States.  The report divides the Canadian analysis in two main sections, presenting the Canadian patent landscape from the perspective of Canadian institutions and also of Canadian researchers. These two sections provide a detailed overview of filing activity looking at areas of specialization by AI sub-category, key players, geographical distribution across the country, and patent landscape maps. The details presented in these sections are useful to better understand the evolution and the current state of innovation in this technology field. The report showcases a number of complementary research activities undertaken including as section that derives from CIPO’s collaboration with Statistics Canada, as well as the latest metric developed at CIPO called the IP Concentration Index (IPCI).

 This case study provides an example of how to know better patent applicants at the French Patent Office using the French patent database.

Key questions for this applicants’ analysis are:

• What are the different types of patent applicants?
• What are the geographical localisations of patent applicants?
•  What are the technological domains of applicants and do they have several domains?
•  Do patent applicants often extend their French patent applications?
•  What are the links between applicants’ types and technological domains, international extension, or geographical localisations?

 Among the fields wherein AI gains more and more importance, bioinformatics is one of the most thrilling and variegated, with applications ranging from diagnostics to genomics or designing personalized medicine.

Its patent landscape is explored by means of Patent Analytics, Semantic search and AI tools in order to identify trends and new players.

 The semantic AI-text-based search methodologies adds a new option to the searcher’s toolbox for helping to explore the prior art. What is the difference to a regular Boolean search? How do I use them and when? Seeming, we don’t have much of instruction manuals teaching us how to use it efficiently. This presentation will take a glimpse under the hood and look into best practices. It will include a review of performance statistics, quality measurements and, last but not least, usage challenges. Finally, some personal thoughts on the future; what to expect next in the field of information retrieval and how the searcher’s role might evolve onwards.

After having successfully implemented a proof of concept in late 2019, DS9 started with one of its customers to implement a pilot for a Deep Learning based personal News Rating system.
The system was initially trained to rate news matching a limited number of typical user profiles. Over time the system would evolve and learn more and more about personal preferences and build user-specific rating models. 
This talk looks at the many challenges we overcame during implementation: collecting training data, integrating many diverse news sources, management of many user-specific models and automatic deployment to production.

We have created structure based chemical ontologies that are used to classify chemical compounds automatically. These classifications can be used with success in semantic search engines to find all representatives of a chemical class. In the present paper we would like to demonstrate use cases when utilizing these chemical classes as features in typical machine learning approaches.
Thus, we have used the co-occurrence of chemical compounds with biological and physico-chemical properties in scientific articles to train models that predict properties of novel compounds that did not occur in those training sets. One example is the prediction of hepatotoxicity as well as bioavailability. In principle, one can use any property that is found in the textual vicinity of compounds to build such predictive models. Criteria will be presented that allow to judge the quality and predictive power of such models.

With all new technologies and intelligence one may think that all information issues will be solved in the (near) future. However, one of the most fundamental issues at hand is that with the lack of reliable, quality information there is no useable output to work with in the first place. This presentation looks at the global challenges that we are still faced with today relating to content that will keep us from truly intelligent discovery in the future if nothing is done.

Patent landscape efforts can get hampered either by voluminous patent search results or the perceived need to manually tag every single feature. It can increase uncertainty, costs, and complexity. Like chemical structure, biosequence, or freedom-to-operate patent searches, patent landscape searches have unique challenges. Delivering custom patent landscape analysis for effective decisions is beyond the skills of most end users. The final patent landscape report can vary depending on the allocated time and the optimal usage of advanced patent analytics tools by professional patent analysts (or by end-users). Many “state of the art” reports only provide patent search results with minimal analysis. Similarly, many automated “instant reports” only provide canned analysis and visuals that are too broad for POV analysis or corporate decision-making. This presentation describes the use of an efficient accelerated, intentional, and multifaceted (AIM™) patent landscape to alleviate these issues. The goal of an AIM™ patent landscape is to intentionally align scope with pending decisions. The results are delivered with appropriate level of analysis, including supporting charts, within an accelerated timeline of 2-3 weeks. AIM™ patent landscape analysis uses experienced patent analysts, a well-defined workflow process, and multiple best-in class patent data search, processing, and visualization tools. AIM™ patent landscapes are ideal in patent portfolio benchmarking efforts and delineating white space opportunities for well-defined projects.

Discover how an AI-based question-answering system can combine semantic modelling, ontologies, linguistics and artificial intelligence algorithms in a self-refining system that delivers results based on inter-related meaning of facts. Structured and non-structured data are integrated in a system that not only allows for phrase searches and structured queries but also offers its users a unique hybrid natural language question answering system. Machine learning algorithms are combined with semantic network-based "prior knowledge" inference. The system integrates seamlessly with existing infrastructure and does not just integrate a static, public "world graph" of common knowledge, but builds "digital trust" by offering organization-specific background information and helps leverage knowledge buried both in decade old data as well as data derived from news feeds providing real-time semantic analysis of breaking news.

In the last twelve months AI activity has continued to accelerate.  While there have been major setbacks in AI over the decades its recent up surge seems to be holding.  Many positives stories are hitting the news, but is anyone actually making any money on AI deployments besides the big AI vendors?  Have there been significant, meaningful cost reductions from AI deployments?  Yes!  Brief case studies will be presented from primary and secondary sources illustrating impacts on real world cost savings and revenue enhancements.  As is always the case with real world projects there are lessons learned!