Automatic Patent Classification Research Articles

Forty Years of Automated Patent Classification

The continuous development of technology increases the number of new inventions and thus the number of patents. Each new technology is assigned to one of the existing patent classes, otherwise, a new technology must be added to the patent classes. These processes support the development of automated patent classification methods and articles on this subject. For decades, studies have been conducted in the field of automatic patent classification. In this paper, the studies and pat ents published in this field have been evaluated using a bibliometric approach. The most referenced important studies on this subject are examined in detail, the analysis is detailed by making the pioneering authors of the field, journals, countries, assignees, etc., the focal point. The most recent developments in the field have been examined and evaluations have been presented regarding topics that are likely to be studied in the future.

Reliable Multi-View Deep Patent Classification

Patent classification has long been regarded as a crucial task in patent information management and patent knowledge mining. In recent years, studies combining deep learning automatic patent classification methods with deep neural networks have significantly increased. Although great efforts have been made in the patent deep classification task, they mainly focus on information extraction from a single view (e.g., title or abstract view), but few studies concern multi-view deep patent classification, which aims to improve patent classification performance by integrating information from different views. To that end, we propose a reliable multi-view deep patent classification method. Within this method, we fuse multi-view patent information at the evidence level from the perspective of evidence theory, which not only effectively improves classification performance but also provides a reliable uncertainty estimation to solve the unreliability of classification results caused by property differences and inconsistencies in the different patent information sources. In addition, we theoretically prove that our approach can reduce the uncertainty of classification results through the fusion of multiple patent views, thus facilitating the performance and reliability of the classification results. The experimental results on 759,809 real-world multi-view patent data in Shanghai, China, demonstrate the effectiveness, reliability, and robustness of our approach.

Deep learning based pipeline with multichannel inputs for patent classification

This work introduces a deep learning pipeline for automatic patent classification with multichannel inputs based on LSTM and word vector embeddings. Sophisticated text mining methods are used to extract the most important segments from patent texts, and a domain-specific pre-trained word embeddings model for the patent domain is developed; it was trained on a very large dataset of more than five million patents. The deep learning pipeline is using multiple parallel LSTM networks that read the source patent document using different input dimensions namely embeddings of different segments of patent texts, and sparse linear input of different metadata. Classifying patents into corresponding technical fields is selected as a use case. In this use case, a series of patent classification experiments are conducted on different patent datasets, and the experimental results indicate that using the segments of patent texts as well as the metadata as multichannel inputs for a deep neural network model, achieves better performance than one input channel.

PERBANDINGAN KINERJA METODE PRA-PEMROSESAN DALAM PENGKLASIFIKASIAN OTOMATIS DOKUMEN PATEN

This paper presents a performance analysis and comparison of several pre-processing methods used in automatic patent classification with graph kernels for Support Vector Machine (SVM). The pre-processing methods are based on the data transform techniques, namely data scaling, data centering, data standardization, data normalization, the Box-Cox transform and the Yeo-Johnson transform. The automatic patent classification is designed to classify an input of patent citation graphs into one of 10 possible classes of the International Patent Classification (IPC). The input is taken with various background conditions. The experiments showed that the best result is achieved when the pre-processing method is data normalization, achieving a classification accuracy of up to 85.33.15% for the KEHL and 93.80% for the KVHL. In contrast, for the KEHG, the preprocessing method application decreased the accuracy.

Automated classification of patents: A topic modeling approach

Due to the rapid increase in technological innovation and corresponding increase in patent applications, automatic patent classification systems are very helpful for both individual inventors and patent attorneys in classifying patents. However, previous studies have neglected the question of what content patents include and how to represent patent content effectively in a structured form to predict the patent class. In response, this study suggests a topic model based on support vector machine (SVM) prediction for automatic patent classification. This study considers two important issues for patent classification: text representation and class prediction. For text representation, we use the topic modeling technique and employ latent Dirichlet allocation (LDA). The result of LDA is then used as the input for the second aspect: class prediction. We use SVM prediction for automatic patent classification. We also suggest potential improvement strategies to enhance the prediction performance of our suggested approach. This study contributes to the field in that it can lead to the automatic classification of patents without the need for any expert judgment during the process.

Domain-specific word embeddings for patent classification

PurposePatent offices and other stakeholders in the patent domain need to classify patent applications according to a standardized classification scheme. The purpose of this paper is to examine the novelty of an application it can then be compared to previously granted patents in the same class. Automatic classification would be highly beneficial, because of the large volume of patents and the domain-specific knowledge needed to accomplish this costly manual task. However, a challenge for the automation is patent-specific language use, such as special vocabulary and phrases.Design/methodology/approachTo account for this language use, the authors present domain-specific pre-trained word embeddings for the patent domain. The authors train the model on a very large data set of more than 5m patents and evaluate it at the task of patent classification. To this end, the authors propose a deep learning approach based on gated recurrent units for automatic patent classification built on the trained word embeddings.FindingsExperiments on a standardized evaluation data set show that the approach increases average precision for patent classification by 17 percent compared to state-of-the-art approaches. In this paper, the authors further investigate the model’s strengths and weaknesses. An extensive error analysis reveals that the learned embeddings indeed mirror patent-specific language use. The imbalanced training data and underrepresented classes are the most difficult remaining challenge.Originality/valueThe proposed approach fulfills the need for domain-specific word embeddings for downstream tasks in the patent domain, such as patent classification or patent analysis.

Combined graph kernels for automatic patent classification: A hybrid approach

In this study, we proposed combined kernel-based methods to leverage patent citation graph performance for patent classification. The concept is to use the combined graph kernels of the citation graph to classify patent documents, as a hybrid approach. A multiple kernel framework was used for integrating multiple datasets of various kernels into a combined kernel. We employed seven graph kernels as the baselines and the combination of random walks and Weisfeiler–Lehman subtree kernels to achieve higher performance. We calculated the kernel values of each patent pairwise and employed an SVM classifier to carry out the classification task. The investigation results demonstrate that the combined graph kernel outperforms single kernels.

A Hierarchical Feature Extraction Model for Multi-Label Mechanical Patent Classification

Various studies have focused on feature extraction methods for automatic patent classification in recent years. However, most of these approaches are based on the knowledge from experts in related domains. Here we propose a hierarchical feature extraction model (HFEM) for multi-label mechanical patent classification, which is able to capture both local features of phrases as well as global and temporal semantics. First, a n-gram feature extractor based on convolutional neural networks (CNNs) is designed to extract salient local lexical-level features. Next, a long dependency feature extraction model based on the bidirectional long–short-term memory (BiLSTM) neural network model is proposed to capture sequential correlations from higher-level sequence representations. Then the HFEM algorithm and its hierarchical feature extraction architecture are detailed. We establish the training, validation and test datasets, containing 72,532, 18,133, and 2679 mechanical patent documents, respectively, and then check the performance of HFEMs. Finally, we compared the results of the proposed HFEM and three other single neural network models, namely CNN, long–short-term memory (LSTM), and BiLSTM. The experimental results indicate that our proposed HFEM outperforms the other compared models in both precision and recall.

Product functional information based automatic patent classification: Method and experimental studies

In order to effectively extract the hidden information from the patent texts and to further provide this information to support the product innovation design process, this paper proposed an automatic patent classification method based on the functional basis and Naive Bayes theory. The functions of products are regarded as the innovation attributes, and the function co-reference relations of the patents in different areas are established. Patent classification methods are proposed based on the functions of products and the general steps of the patent classification process are proposed. In addition, three training methods are studied in the experiments, including multi-classification fully supervised training, multiple dichotomous supervised training and semi-supervised training. Through comparing and analyzing the experimental results, a patent text classifier is developed. In summary, this paper provides a general idea and the relevant technologies on how to build a patent knowledge space by automatically extracting and expanding the patent texts.

A patent quality classification model based on an artificial immune system

Patents are business and financial assets which can enhance a company's competitive position. Thus, patent analysis is important for defining business strategies and supporting decision-making in organizations. However, patent analysis can involve vast data sets and are difficult to analyze. The purpose of this study is to apply artificial immune system hybrid collaborative filtering to build a patent quality classification model. We apply the model to predicting the quality of radio frequency identification patents. Using a simple definition of quality, we define each patent's data as an antigen and then compute the affinities of the target patent to all immune networks. If the affinity is larger than a given threshold, the antibody is cloned to the related immune network. After the immune networks are constructed, they exhibit high affinity to the target patent. Finally, a series of experiments show that the proposed model can accurately predict the quality of new patents. The resulting automatic patent quality classification model provides manufacturers with improved insights into their company's intellectual property strategy, product direction and long-term vision.

A patent quality analysis and classification system using self-organizing maps with support vector machine

A plethora of patents are approved by the patent officers each year and current patent systems face a solemn quandary of evaluating these patents’ qualities. Traditional researchers and analyzers have fixated on developing sundry patent quality indicators only, but these indicators do not have further prognosticating power on incipient patent applications or publications. Therefore, the data mining (DM) approaches are employed in this article to identify and to classify the new patent's quality in time. An automatic patent quality analysis and classification system, namely SOM-KPCA-SVM, is developed according to patent quality indicators and characteristics, respectively. First, the self-organizing map (SOM) approach is used to cluster patents published before into different quality groups according to the patent quality indicators and defines group quality type instead of via experts. The kernel principal component analysis (KPCA) approach is used to transform nonlinear feature space in order to improve classification performance. Finally, the support vector machine (SVM) is used to build up the patent quality classification model. The proposed SOM-KPCA-SVM is applied to classify patent quality automatically in patent data of the thin film solar cell. Experimental results show that our proposed system can capture the analysis effectively compared with traditional manpower approach.

A Supervised Requirement-oriented Patent Classification Scheme Based on the Combination of Metadata and Citation Information

AbstractPatent classification systems are applied extensively in innovative analysis. Existing patent classification schemes are either technology-dependent or TRIZ-based. The former ones, such as the IPC and UPC, are normally developed by different patent offices in the world mainly for the purpose of patentability examination and patent retrieval, while the latter is for TRIZ users and analysts with no more than 40 categories. These static classifications are too complex and general to meet the in-depth patent classification requirements of a specific technology area or organization. To tackle these drawbacks, in this paper, we propose an automatic requirement-oriented patent classification scheme as a complementary method using supervised machine learning techniques to classify patent dataset into a user-defined taxonomy. The requirement-oriented patent taxonomy can be technology-dependent, application-dependent or a mixture of both tailored to specific business objectives. It is more comprehensible an...

Grouping of TRIZ Inventive Principles to facilitate automatic patent classification

Automatic patent classification facilitates searching for previous patent documents. For TRIZ users, they would like to search for patents based on the solutions (TRIZ Inventive Principles) to the Contradictions addressed in the patents, which is different from traditional searching for prior arts based on the application fields of the inventions. For this purpose, a TRIZ-based patent classification expert system is needed. To facilitate automatic classification of patent documents according to Inventive Principles (IPs) for TRIZ users, we analyze the original 40 IPs proposed by Altshuller. Seven IPs are defined as Obscure IPs, the other 33 as Distinct IPs. Furthermore, two kinds of similarity among the Distinct IPs are defined: text similarity and meaning similarity. Then the 40 IPs are grouped into 22 new classes. Automatic classification based on 674 patent documents associated with these 22 new classes is tested and analyzed, with two issues of multi-label classification and class imbalance addressed.

Automatic classification of patent documents for TRIZ users

In contrast to traditional inventors, inventors using TRIZ are not only interested in searching for prior art in related fields, but also for the analogous inventions in other fields that have solved the same Technical Contradiction by using the same method. To be useful for TRIZ users, patents are required to be classified by the Contradiction they solved and Inventive Principles they used instead of the fields in which they are involved. Most of the currently available automatic patent classification systems are based on technology-dependent schemes such as the IPC and they cannot satisfy TRIZ users’ requirements. In this paper, an automatic patent classification for TRIZ users is proposed and explained in detail. In a preliminary study, patent documents were collected for 6 out of 40 Inventive Principles, and the proposed automatic classification tested.