Extraction of the effect and the technology terms from a patent document

Abstract

A patent map, a visual representation of related patent information, is an effective strategic tool for analysis of patent application trends. In particular, an effect-technology type patent map is commonly used as patent examinations are based on technology and effect terms of the inventions.However, conventional patent mining tools are not realized to make an effect-technology type patent map.In order to generate effect-technology type patent maps automatically, we propose a method for extracting information on the effect and the technological terms from patent documents. Our proposed method is described by the following steps. (Step 1) Extract expressions containing the information on the technical effect by using frequent expressions and clue expressions that are automatically extracted by using statistical information and initial clue expressions. (Step 2) Extract technology terms in the above expressions by character-string comparison between patent claims and the expressions. (Step 3) Extract effect terms using grammar patterns that are composed of technology terms, frequent expressions and clue expressions.Our method achieves high precision (85.0%) by using frequent expressions, clue expressions, and the grammar patterns.

Extracting characteristic terms from patent documents

“…Next, we select the features used to generate the deep learning model. Here, the proposed method assigns weight ( , ) to content words (nouns, verbs, and adjectives) contained in positive examples in the training data using Equation (1).…”

“…However, according to statistics issued by the patent office, more than 300,000 patent applications have been submitted each year over the past 10 years (2008 to 2017). Analyzing such an enormous number of patent documents requires a significant manual labor: therefore, interest in tools that assist patent analysis has increased, and various patent analysis methods have been developed to analyze large quantities of patent documents automatically [1][2] [3].…”

Development of an entropy-based feature selection method and analysis of online reviews on real estate

“…A positive keyword is determined when ( 7) is true, and likewise, a negative keyword is determined when ( 8) is true. (7) (8)…”

“…He et.al., [6] analyzed web marketing of pizza industry by text-mining tools applying to Tweets and Facebook data. Nonaka et al, [7,8] extracted technology words and user needs from patent documents by using an entropy and grammatical pattern based method.…”

Community detection and growth potential prediction from patent citation networks

“…A clue phrase was defined as the word linking a technology term and an effect term. Sakai et al [7] proposed an automatic collection of clue phrases from patent documents by using an entropy-based score, and Nonaka et al [8] expanded it to extract information on the effect and the technological terms from patent documents by using grammar patterns.…”