The state-of-the-art on Intellectual Property Analytics (IPA): A literature review on artificial intelligence, machine learning and deep learning methods for analysing intellectual property (IP) data

Optical MRI imaging based on computer vision for extracting and analyzing morphological features of renal tumors

Computer vision technology is more and more widely used in the market. Target detection and feature extraction are two important auxiliary means of this technique, which are helpful to analyze target motion data. However, in the field of biology, there are some data limitations in the analysis of targets such as bacteria and tumors, which need to be further explored. Optical MRI imaging technology based on computer vision provides a new way to extract and analyze morphological features of renal tumors. In this paper, an optical MRI imaging method based on computer vision is designed and developed for the extraction and analysis of morphological features of kidney tumors. By using optical MRI imaging technology based on computer vision, the morphological characteristics of kidney tumors were extracted by analyzing the optical characteristics and MRI images of kidney tumors, and a simulation model was established to simulate the morphological characteristics of different types of kidney tumors, and feature extraction and analysis were carried out by computer algorithm. Through the analysis of the simulation model, the morphological characteristics of renal tumors were extracted and analyzed, which provided a new and non-invasive method for clinical diagnosis and treatment of renal tumors.

Tech mining: a revisit and navigation

This mini-review arrays the pertinent tools and purposes of "Tech Mining" - shorthand for empirical analyses of Science, Technology and Innovation (ST&I) data. The intent is to introduce the range of tools, and show how they can complement each other. Tech Mining aims to generate powerful intelligence to help manage R&D and innovation processes. We offer a 5-part array to help relate the analytical elements. An overview of a case study of Hybrid and Electric Vehicles illustrates the complexities involved and the potential to generate valuable "intel."

Bayesian inference of spatial and temporal relations in AI patents for EU countries

In the paper, we propose two models of Artificial Intelligence (AI) patents in European Union (EU) countries addressing spatial and temporal behaviour. In particular, the models can quantitatively describe the interaction between countries or explain the rapidly growing trends in AI patents. For spatial analysis Poisson regression is used to explain collaboration between a pair of countries measured by the number of common patents. Through Bayesian inference, we estimated the strengths of interactions between countries in the EU and the rest of the world. In particular, a significant lack of cooperation has been identified for some pairs of countries. Alternatively, an inhomogeneous Poisson process combined with the logistic curve growth accurately models the temporal behaviour by an accurate trend line. Bayesian analysis in the time domain revealed an upcoming slowdown in patenting intensity.

iPromoter-CLA: Identifying promoters and their strength by deep capsule networks with bidirectional long short-term memory

BACKGROUND AND OBJECTIVE

The promoter is a fragment of DNA and a specific sequence with transcriptional regulation function in DNA. Promoters are located upstream at the transcription start site, which is used to initiate downstream gene expression. So far, promoter identification is mainly achieved by biological methods, which often require more effort. It has become a more effective classification and prediction method to identify promoter types through computational methods.

METHODS

In this study, we proposed a new capsule network and recurrent neural network hybrid model to identify promoters and predict their strength. Firstly, we used one-hot to encode DNA sequence. Secondly, we used three one-dimensional convolutional layers, a one-dimensional convolutional capsule layer and digit capsule layer to learn local features. Thirdly, a bidirectional long short-time memory was utilized to extract global features. Finally, we adopted the self-attention mechanism to improve the contribution of relatively important features, which further enhances the performance of the model.

RESULTS

Our model attains a cross-validation accuracy of 86% and 73.46% in prokaryotic promoter recognition and their strength prediction, which showcases a better performance compared with the existing approaches in both the first layer promoter identification and the second layer promoter's strength prediction.

CONCLUSIONS

our model not only combines convolutional neural network and capsule layer but also uses a self-attention mechanism to better capture hidden information features from the perspective of sequence. Thus, we hope that our model can be widely applied to other components.

Crisis-Critical Intellectual Property: Findings From the COVID-19 Pandemic

A pandemic calls for large-scale action across national and international innovation systems in order to mobilize resources for developing and manufacturing crisis-critical products efficiently and in the huge quantities needed. Nowadays, these products also include a wide range of digital innovations. Given that many responses to the pandemic are technology driven, stakeholders involved in the development and manufacturing of crisis-critical products are likely to face intellectual property (IP)-related challenges. To (governmental) decision makers, IP challenges might not appear to be of paramount urgency compared to the many undoubtedly huge operational challenges to deploy critical resources. However, if IP challenges are considered too late, they may cause delays to urgently mobilize resources effectively. Innovation stakeholders could then be reluctant to fully engage in the development and manufacturing of crisis-critical products. This article adopts an IP and innovation perspective to learn from the currently unfolding COVID-19 pandemic using secondary data, including patent data, synthesized with an IP roadmap. We focus on technical aspects related to research, development, and upscaling of capacity to manufacture crisis-critical products in the huge volumes suddenly in demand. In this article, we offer a set of contributions. We provide a structure, framework, and language for those concerned with steering clear of IP challenges to avoid delays in fighting a pandemic. We provide a reasoning why IP needs to be considered earlier rather than too late in a global health crisis. Major stakeholders we identify include 1) governments; 2) manufacturing firms owning existing crisis-critical IP (incumbents in crisis-critical sectors); 3) manufacturing firms normally not producing crisis-critical products suddenly rushing into crisis-critical sectors to support the manufacturing of crisis-critical products in the quantities that far exceed incumbents' production capacities; and 4) voluntary grassroot initiatives that form during a pandemic, often by highly skilled engineers and scientists in order to contribute to the development and dissemination of crisis-critical products. For these major stakeholders, we draw up three scenarios, from which we identify associated IP challenges they face related to the development and manufacturing of technologies and products for 1) prevention (of spread); 2) diagnosis of infected patients; and 3) the development of treatments. This article provides a terminology to help policy and other decision makers to discuss IP considerations during pandemics. We propose a framework that visualizes changing industrial organizations and IP-associated challenges during a pandemic and derive initial principles to guide innovation and IP policy making during a pandemic. Obviously, our findings result only from observations of one ongoing pandemic and thus need to be verified further and interpreted with care.

Hydrothermal biomass processing for green energy transition: insights derived from principal component analysis of international patents

As efforts to achieve Net Zero are intensifying, there is a strong need to identify the technological positioning of green process innovations that can support the green energy transition. A veritable contender to support these efforts is the hydrothermal biomass processing technology. This process innovation comprises diverse techniques that can convert biomass substrates into valuable low-carbon fuels. Coordination across all available conversion approaches is encouraged to propel the application of those that consider the environmental and sustainability impacts. We assessed the innovation intensity for different techniques under this green process innovation through applying natural language processing and deployment of principal component analysis on patent data. We positioned our techniques within four distinctive groups (intense, dormant, emerging, and exploratory). In this way, we tracked which hydrothermal technique currently dominates international applications and which ones are gaining traction in the future.

•

Innovation intensities for different green innovation techniques were measured.

•

Policymakers have decisive role in advancing techniques at dominant design phase.

•

Emerging technologies aim attaining synthetic natural gas and C5 sugars.

•

Exploratory techniques focus on sewage sludge and connectivity to wastewater plants.

•

Trending techniques point towards achieving a circular economy.

Is handling unbalanced datasets for machine learning uplifts system performance?: A case of diabetic prediction

BACKGROUND AND AIMS

Healthcare is a sensitive sector, and addressing the class imbalance in the healthcare domain is a time-consuming task for machine learning-based systems due to the vast amount of data. This study looks into the impact of socioeconomic disparities on the healthcare data of diabetic patients to make accurate disease predictions.

METHODS

This study proposed a systematic approach of Closest Distance Ranking and Principal Component Analysis to deal with the unbalanced dataset. A typical machine learning technique was used to analyze the proposed approach. The data set of pregnant diabetic women is analysed for accurate detection.

RESULTS

The results of the case are analysed using sensitivity, which demonstrates that the minority class's lack of information makes it impossible to forecast the results. On the other hand, the unbalanced dataset was treated using the proposed technique and evaluated with the machine learning algorithm which significantly increased the performance of the system.

CONCLUSION

The performance of the machine learning-based system was significantly enhanced by the unbalanced dataset which was processed with the proposed technique and evaluated with the machine learning algorithm. For the first time, an unbalanced dataset was treated with a combination of Closest Distance Ranking and Principal Component Analysis.

A Use Case of Patent Classification Using Deep Learning with Transfer Learning

Purpose

Patent classification is one of the areas in Intellectual Property Analytics (IPA), and a growing use case since the number of patent applications has been increasing worldwide. We propose using machine learning algorithms to classify Portuguese patents and evaluate the performance of transfer learning methodologies to solve this task.

Design/methodology/approach

We applied three different approaches in this paper. First, we used a dataset available by INPI to explore traditional machine learning algorithms and ensemble methods. After preprocessing data by applying TF-IDF, FastText and Doc2Vec, the models were evaluated by cross-validation in 5 folds. In a second approach, we used two different Neural Networks architectures, a Convolutional Neural Network (CNN) and a bi-directional Long Short-Term Memory (BiLSTM). Finally, we used pre-trained BERT, DistilBERT, and ULMFiT models in the third approach.

Findings

BERTTimbau, a BERT architecture model pre-trained on a large Portuguese corpus, presented the best results for the task, even though with a performance of only 4% superior to a LinearSVC model using TF-IDF feature engineering.

Research limitations

The dataset was highly imbalanced, as usual in patent applications, so the classes with the lowest samples were expected to present the worst performance. That result happened in some cases, especially in classes with less than 60 training samples.

Practical implications

Patent classification is challenging because of the hierarchical classification system, the context overlap, and the underrepresentation of the classes. However, the final model presented an acceptable performance given the size of the dataset and the task complexity. This model can support the decision and improve the time by proposing a category in the second level of ICP, which is one of the critical phases of the grant patent process.

Originality/value

To our knowledge, the proposed models were never implemented for Portuguese patent classification.

Network Characteristic Control of Social Dilemmas in a Public Good Game: Numerical Simulation of Agent-Based Nonlinear Dynamics

This paper proposes a possible mechanism for obtaining sizeable behavioral structures by simulating a network–agent dynamic on an evolutionary public good game with available social .learning. The model considers a population with a fixed number of players. In each round, the chosen players may contribute part of their value to a common pool. Then, each player may imitate the strategy of another player based on relative payoffs (whoever has the lower payoff adopts the strategy of the other player) and change his or her strategy using different exploratory variables. Relative payoffs are subject to incentives, including participation costs, but may also be subject to mutation, whose rate is sensitized by the network characteristics (social ties). The process discussed in this report is interesting and relevant across a broad range of disciplines that use game theory, including cultural evolutionary dynamics.

Exploring the Open COVID Pledge in the fight against COVID‐19: a semantic analysis of the Manifesto, the pledgors and the featured patents

Coronavirus disease‐19 (COVID‐19) has stimulated urgent innovative responses to tackle the current crisis and unveil new trajectories enabling recovery as early as possible. In the quest for solutions to the pandemic, organizations have been forced to join efforts with an unprecedented number of different stakeholders, including competitors, rising new appropriation‐related challenges. To ease these issues and facilitate collaborative efforts, some initiatives have come into being to encourage the release of Intellectual Property (IP) rights to unlock new possibilities from their use and possibly foster the collective innovation process. The Open COVID Pledge (OCP) stands out as the most visible project that has gained momentum at the international level, as it has increasingly involved well‐known top‐patenting companies, willing to publicly commit to making their IP relevant to COVID‐19 freely available. Drawing from all the available information (the World Wide Web, the participating companies' press releases and official websites, and the documents of pledged patents), we propose a research design, applying a semantic method to allow an augmented understanding of the main characteristics of this pledge. Our findings point out that the OCP has got a great media resonance on the overall web, also thanks to the commitment of large top‐patenting pledgors; results also show that while the official communications of the participant companies resemble very much the general OCP Manifesto of providing free access to their patent portfolio, the semantic analysis of the pledged patents unveils details on available technologies that mostly refer to the real‐time search and analysis of information and devices for the detection of the diffusion of the virus. Overall, this analysis contributes to providing contextual information on the available IP, towards the desired direction of putting the pledge to work and have an impact on follow‐on innovation, which represents the underlying rationale of the initiative in the fight against COVID‐19.

Knowledge Management and Innovation Performance a Mediated-Moderation Model

This study aims at explaining innovation performance of organizations as a combination of resources and capabilities. This study starts with asking the question how the relationship between firm-specific knowledge and innovation performance can be explained. To answer this question, insights from the resource-based view (RBV) and the dynamic capabilities approach (DCA) are combined. This leads to a set of hypotheses. The first hypothesis states that knowledge-specificity and innovation performance are positively related. The second hypothesis states that organizational learning practices mediate the relationship between knowledge-specificity and innovation performance. Then, two contrasting hypotheses are formulated stating that the relationship between knowledge-specificity and organizational learning practices of organizations is strengthened or weakened by the level of autonomy. Together these hypotheses lead to a mediated-moderation model of knowledge-specificity and innovation performance. The model is tested using a mediated-moderation analysis on a sample of 673 private organizations in the Netherlands. The analyses show that there is a positive relationship between knowledge-specificity and innovation performance and that this relationship is mediated by the extent to which organizations apply learning practices. Hypotheses 1 and 2 are thus supported. Furthermore, the level of autonomy weakens the relationship between knowledge-specificity and organizational learning practices. This study’s main contribution lies in combining theoretical insights from the RBV and the DCA, applying them to the field of knowledge management, and testing them empirically. The analyses lead to two insights for organizations interested in increasing their innovation performance. First, investing in learning capabilities enhances innovation performance. Second, organizations based on general knowledge can grant work autonomy to employees to enhance their ability to learn.

Comparison of machine learning methods for estimating case fatality ratios: An Ebola outbreak simulation study

BACKGROUND

Machine learning (ML) algorithms are now increasingly used in infectious disease epidemiology. Epidemiologists should understand how ML algorithms behave within the context of outbreak data where missingness of data is almost ubiquitous.

METHODS

Using simulated data, we use a ML algorithmic framework to evaluate data imputation performance and the resulting case fatality ratio (CFR) estimates, focusing on the scale and type of data missingness (i.e., missing completely at random-MCAR, missing at random-MAR, or missing not at random-MNAR).

RESULTS

Across ML methods, dataset sizes and proportions of training data used, the area under the receiver operating characteristic curve decreased by 7% (median, range: 1%-16%) when missingness was increased from 10% to 40%. Overall reduction in CFR bias for MAR across methods, proportion of missingness, outbreak size and proportion of training data was 0.5% (median, range: 0%-11%).

CONCLUSION

ML methods could reduce bias and increase the precision in CFR estimates at low levels of missingness. However, no method is robust to high percentages of missingness. Thus, a datacentric approach is recommended in outbreak settings-patient survival outcome data should be prioritised for collection and random-sample follow-ups should be implemented to ascertain missing outcomes.

Computational identification of eukaryotic promoters based on cascaded deep capsule neural networks

A promoter is a region in the DNA sequence that defines where the transcription of a gene by RNA polymerase initiates, which is typically located proximal to the transcription start site (TSS). How to correctly identify the gene TSS and the core promoter is essential for our understanding of the transcriptional regulation of genes. As a complement to conventional experimental methods, computational techniques with easy-to-use platforms as essential bioinformatics tools can be effectively applied to annotate the functions and physiological roles of promoters. In this work, we propose a deep learning-based method termed Depicter (Deep learning for predicting promoter), for identifying three specific types of promoters, i.e. promoter sequences with the TATA-box (TATA model), promoter sequences without the TATA-box (non-TATA model), and indistinguishable promoters (TATA and non-TATA model). Depicter is developed based on an up-to-date, species-specific dataset which includes Homo sapiens, Mus musculus, Drosophila melanogaster and Arabidopsis thaliana promoters. A convolutional neural network coupled with capsule layers is proposed to train and optimize the prediction model of Depicter. Extensive benchmarking and independent tests demonstrate that Depicter achieves an improved predictive performance compared with several state-of-the-art methods. The webserver of Depicter is implemented and freely accessible at https://depicter.erc.monash.edu/.

Data science and its relationship to library and information science: a content analysis

Purpose

The purpose of this paper is to present the results of a study exploring the emerging field of data science from the library and information science (LIS) perspective.

Design/methodology/approach

Content analysis of research publications on data science was made of papers published in the Web of Science database to identify the main themes discussed in the publications from the LIS perspective.

Findings

A content analysis of 80 publications is presented. The articles belonged to the six broad categories: data science education and training; knowledge and skills of the data professional; the role of libraries and librarians in the data science movement; tools, techniques and applications of data science; data science from the knowledge management perspective; and data science from the perspective of health sciences. The category of tools, techniques and applications of data science was most addressed by the authors, followed by data science from the perspective of health sciences, data science education and training and knowledge and skills of the data professional. However, several publications fell into several categories because these topics were closely related.

Research limitations/implications

Only publication recorded in the Web of Science database and with the term “data science” in the topic area were analyzed. Therefore, several relevant studies are not discussed in this paper that either were related to other keywords such as “e-science”, “e-research”, “data service”, “data curation”, “research data management” or “scientific data management” or were not present in the Web of Science database.

Originality/value

The paper provides the first exploration by content analysis of the field of data science from the perspective of the LIS.

Use of the Industrial Property System in Colombia (2018): A Supervised Learning Application

The purpose of this paper is to establish ways to predict the spatial distribution of the use of the intellectual property system from information on industrial property applications and grants (distinctive signs and new creations) and copyright registrations in 2018. This will be done using supervised learning algorithms applied to information on industrial property applications and grants (trademarks and new creations) and copyright registrations in 2018. Within the findings, 4 algorithms were identified with a level of explanation higher than 80%: (i) Linear Regression, with an elastic network regularization; (ii) Stochastic Gradient Descent, with Hinge loss function, Ringe regularization (L2) and a constant learning rate; (iii) Neural Networks, with 1,000 layers, with Adam’s solution algorithm and 2,000 iterations; (iv) Random Forest, with 10 trees.

International collaboration generates high quality clusters of pharmaceutical patents

The worldwide active patent portfolio has nearly doubled in numbers and strength since 2000. The number of active pharmaceutical patent families has tripled in the same time period. The quantitative growth results mostly from a surge of patents from China, half of them classified in A61K36 ('medicinal preparations of undetermined constitution containing material from algae, lichens, fungi or plants'). High-quality patents exhibit a slower growth curve, and cluster within the three areas biologicals; heterocyclic compounds, and cancer drugs. However, the highest concentration of high-quality patents was found when selecting patents listing inventors from at least two out of the five most important countries of origin for pharmaceutical patents: China, EP countries, Japan, South Korea and the USA.

Data science from a library and information science perspective

Purpose

Data science is a relatively new field which has gained considerable attention in recent years. This new field requires a wide range of knowledge and skills from different disciplines including mathematics and statistics, computer science and information science. The purpose of this paper is to present the results of the study that explored the field of data science from the library and information science (LIS) perspective.

Design/methodology/approach

Analysis of research publications on data science was made on the basis of papers published in the Web of Science database. The following research questions were proposed: What are the main tendencies in publication years, document types, countries of origin, source titles, authors of publications, affiliations of the article authors and the most cited articles related to data science in the field of LIS? What are the main themes discussed in the publications from the LIS perspective?

Findings

The highest contribution to data science comes from the computer science research community. The contribution of information science and library science community is quite small. However, there has been continuous increase in articles from the year 2015. The main document types are journal articles, followed by conference proceedings and editorial material. The top three journals that publish data science papers from the LIS perspective are the Journal of the American Medical Informatics Association, the International Journal of Information Management and the Journal of the Association for Information Science and Technology. The top five countries publishing are USA, China, England, Australia and India. The most cited article has got 112 citations. The analysis revealed that the data science field is quite interdisciplinary by nature. In addition to the field of LIS the papers belonged to several other research areas. The reviewed articles belonged to the six broad categories: data science education and training; knowledge and skills of the data professional; the role of libraries and librarians in the data science movement; tools, techniques and applications of data science; data science from the knowledge management perspective; and data science from the perspective of health sciences.

Research limitations/implications

The limitations of this research are that this study only analyzed research papers in the Web of Science database and therefore only covers a certain amount of scientific papers published in the field of LIS. In addition, only publications with the term “data science” in the topic area of the Web of Science database were analyzed. Therefore, several relevant studies are not discussed in this paper that are not reflected in the Web of Science database or were related to other keywords such as “e-science,” “e-research,” “data service,” “data curation” or “research data management.”

Originality/value

The field of data science has not been explored using bibliographic analysis of publications from the perspective of the LIS. This paper helps to better understand the field of data science and the perspectives for information professionals.