IP Innovation Trends & Future Directions

Predicting industrial development trends in a target area using publicly available data like enterprise addresses, policy support, and disclosures. The method involves identifying high-quality enterprises in the area based on published data, classifying them by industry, and analyzing their size and intellectual property to predict industry trends. This leverages artificial intelligence techniques like support vector machines and gradient boosting trees to analyze publicly available data and identify advantageous industrial chains and trends in the target area.

Source

Method and system for automated patent infringement risk assessment using large-scale natural language models. The method involves generating a technology development tree for a product or process using a natural language model. The tree captures the technical features and concepts at each step of the development process. Patents are then analyzed using the same model to extract their technical features. By comparing the extracted features against the development tree, infringement risks can be identified and prioritized. This automated process reduces manual effort and time required for patent infringement risk assessment, especially for companies with large patent portfolios.

Source

Analyzing technical line changes in a field based on patent data to understand technology development trends, predict future directions, and help companies make informed innovation strategies. The method involves processing patents to extract key entities, relationships, and time points. It then analyzes time series trends, builds roadmaps, assesses future technology, fuses data, predicts trends, and generates comprehensive reports.

Source

A method and system for predicting disruptive patents using intelligent models. The method involves identifying potentially disruptive technology themes from patent data using machine learning algorithms, scoring the themes using a disruptive technology measurement model, and selecting the top 10% with highest scores for further analysis. It combines SVM-LDA, indicator system construction, and cosine similarity algorithms to improve accuracy in identifying disruptive technologies in complex, uncertain environments.

Source

A patent value evaluation method that combines traditional methods and machine learning to provide more accurate and reliable evaluation results for patents. It captures patent information and potential citation relationships through network data retrieval, determines stage-based weights, uses association rule mining, self-learning, and feedback to classify and value patents based on correlation matrices. This provides more comprehensive, dynamic, and real-time patent value evaluation.

Source

Method for analyzing patents to provide recommendations for optimizing technical protection strategies. The method involves preprocessing the patent data using NLP techniques, extracting text features using deep learning, and then generating reports on technical adaptability, potential infringement, and optimized protection strategies. It leverages techniques like statistical analysis, neural networks, and optimization algorithms to comprehensively understand patent technology, market adaptability, and infringement risks.

Source

Using deep learning and natural language processing (NLP) to predict the economic value of patents. The method involves leveraging unstructured patent text in addition to structured patent features to develop more accurate and scalable patent valuation models. It uses deep learning architectures like bidirectional long short-term memory (LSTM) networks to process the patent text and extract insights. The text features are combined with structured patent data like claims, citations, and classifications to train the models. The resulting deep learning models outperform traditional regression-based methods for patent valuation.

Source

A method and device for predicting promising technologies using an autoregressive integrated moving average (ARIMA) model. The ARIMA model is used to identify time series patterns and predict future technology trends. The method involves identifying an appropriate ARIMA model for the time series of technology data, normalizing the data if needed, and using the identified model to make predictions.

Source

Method for analyzing the industrial feasibility of patents using patent data to determine if a patent has practical application potential. The method involves analyzing patents based on industry terms, citations, transfers, and values to train a model that can predict the feasibility of new patents. It determines average metrics like citations and transfers for patents in different stages (inception, development, application) to train a model. This model is then used to analyze new patents and provide feasibility scores based on the metrics.

Source

Method for analyzing the subject content and popularity evolution of patented technologies by leveraging the International Patent Classification (IPC) system. The method involves using a topic modeling technique called Partial Latent Dirichlet Allocation (pLDA) to extract technical topics from patent documents. The pLDA model is modified to take into account the IPC classification level and abstract text of the patent. By setting the IPC level for topic mining, it allows fine-grained analysis of technical topics at different levels. In subject evolution analysis, word clouds are used to show the subject content under IPC classifications over time. The topic intensity and trend are calculated to determine the hotness trend of topics under IPC classifications.

Source

Method for analyzing technology trends using patent data and classification systems to identify detailed technologies and their convergence, and objectively quantify technology development trends. The method involves extracting patent classification systems, configuring a matrix, clustering to form detailed technology groups, identifying technologies within each cluster, and applying time series analysis to track development trends.

Source

Patent valuation method using data mining and heterogeneous knowledge association to accurately measure patent value. The method involves representing patents and external entities in a heterogeneous knowledge network where patents are connected to external entities like market trends. Machine learning is then used to analyze the network to determine patent value based on the interconnectivity between patent and external entity features.

Source

Predicting future innovation at the company level based on big data and predictive analysis using machine learning techniques that explore the usefulness of company financial data, newspaper articles, social media data, and patent indicators.

Source

Efficiently and quickly mining core patents in a target field by leveraging web crawling, document subject extraction, novelty evaluation, and influence optimization techniques. The method involves using a web crawler to extract patent documents, generating subject sets for all patents, calculating novelty based on cited references, determining influence using citations, and optimizing to find core patents in the target field.

Source

Diagnosing and predicting science and technology capabilities of countries and companies using patent and thesis data. The method involves collecting patent and thesis data for a technology, classifying it by country/institution, calculating patent/thesis variables, applying to machine learning models to diagnose tech strength, then using time series analysis to predict future capabilities. The variables could be things like number of patents, citations, or publications. The models could be regression or neural networks.

Source

Extracting promising technologies through big data analysis using LSTM recurrent neural networks. The method involves cyclically analyzing past patent data with LSTM to predict future promising technologies. LSTM improves prediction accuracy by adding cell state values to the hidden state of recurrent neural networks.

Source

Analyzing enterprise technology using patent big data to provide targeted competitor analysis reports. The method involves collecting patent data for each enterprise, analyzing it to extract label data, then customizing analysis based on target enterprise instructions. Data dimensions like invention trends, patent types, and legal status are analyzed to provide insights into technological development, innovation, talent, competition, cooperation, and technology transfer.

Source

Method for accurately and objectively evaluating the value of intellectual property assets like patents to improve management and governance of intellectual property portfolios. The method involves calculating the value of each dimension (scope, market, quality) of the patent data, then determining the first patent value type based on the overall value. Next, rating each dimension separately to get the second value type. Finally, comparing the first and second types to determine the final, more accurate patent value type. This multi-step evaluation provides a more objective and accurate determination of patent value compared to just relying on a single evaluator.

Source

A method for screening high-value patents using big data and AI to efficiently identify patents with potential for converting into valuable innovations. The method involves steps like setting up a patent database, cleaning and transforming the data, training AI models, and evaluating patents based on factors like patent family size, cited patents, litigation, claims, inventors, and innovation teams. This allows rapid screening of large numbers of patents to identify those with potential for significant innovation impact.

Source

A machine learning-based patent evaluation method that provides detailed and quantitative patent value assessment using multiple categories subdivided based on split points. The method involves extracting quantitative information like bibliographic data and transfer fees from patents, embedding the textual data to convert it into numerical form, training patent classification models with the preprocessed data and split points, and using these models to evaluate the value of a target patent based on its embedded text and quantitative data. The models divide the patents into categories based on split points derived from sorting the training data by transfer fees.

Source