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

Abstract

Patent 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 and adaptable to various patent management requirements. Through a set of experiments on a collection of 14,414 patents in a case study in the technology area of system on a chip (SoC), we recommend using the combination of the metadata and citation information as the document representation for the new method since it can obtain relatively high classification accuracy with a dramatically simplified document preprocessing process.

Semantic TRIZ feasibility in technology development, innovation, and production: A systematic review

“… Identified occurrence frequencies of each verb in all As and each noun in all Os, and referred to the terms in the function/attribute database scheme of TRIZ USPTO SAO structure Semantic functional similarity measurement – Semantic Functional Similarity Measurement (Similarity Coefficient) A24. A supervised ML technique used for automating patent classification by applying text mining [ 52 ]. requirement-oriented taxonomies USPTO Tokenization Stop Words Stemming term weighting Information Gain (IG) TF-IDF Vector Space Model (VSM) Decision tree (DT) Naïve Bayes (NB) Support Vector Machine (SVM) Accuracy, Precision, Recall and F-measure A25.…”

“…Another common text vectorization technique is term frequency-inverse document frequency (TF-IDF), which statistically measures the relevance of a word to a document [ 52 ]. In fact, two different metrics are multiplied to obtain the weight of words in a document [ 31 , 67 ].…”

“…Additionally, “term clumping” method by taking advantage of NLP techniques has been utilized in cleaning and clustering large collections of technical text documents such as patents to obtain information and knowledge in a specific technical domain [ 61 , 62 , 71 ]. It integrates numerous NLP techniques such as removing stop words and constructing synonym lists, fuzzy set matching, TF-IDF, and PCA [ 47 , 52 , 56 , 57 , 60 , 61 ].…”

Analysis of the effect of data properties in automated patent classification

“…Some other works have used the WIPO, NTCIR, CLEF-IP or other patent datasets, but they used them as general text/graph datasets to conduct several forms of classification. These works had different goals than the particularities of patent classification, such as testing the efficiency and/or scalability of their particular methods in general text classification , hierarchical classification , or node classification in graphs (Dallachiesa et al 2014); testing methods for extreme machine learning or dimensionality reduction (Shalaby et al 2014); quantifying the existence of concept drift in data (D'hondt et al 2014); or classifying the data in user-defined hierarchies (Zhu et al 2015). Most of these works used only the title, abstract or claims section from patents, used general accuracy and macro and/or micro-F1 as performance metrics, did not mention what patent codes they used, and considered or not the hierarchy depending on the problem they were studying.…”

Research on Patent-Knowledge Representation and Automatic Classification Based on Situation Mapping

“…Ghareb [21], Labani [22], and Chen [23] proposed several methods for feature selection of patent texts, which could effectively support the attribute extraction of patent texts. Zhu [24] proposed an automatic requirement-oriented patent-classification method to better meet various patent-management requirements. Wu [25] proposed an automatic classification method based on selforganizing maps and support vector machine (SVM), which can help in effectively analyzing the quality of a patent.…”