Information Technology Reference
In-Depth Information
little higher than our MCIVC algorithm, its precision is much lower than other algo-
rithms because it usually retrieves too many redundant classes.
The performance measurement we used to evaluate the final performance of
MCIVC is the accuracy. The performance measurement should get scores only when
all predicted Conflicting Engineer Parameters are exactly the same as all real Con-
flicting Engineer Parameters in the testing patent. The experiment results from the
point of view of technical contradiction classification in final phase are shown in
Table 2(b). The algorithm we proposed gets the highest accuracy. Therefore, we can
conclude that our MCIVC algorithm performs better than other algorithms.
5
Conclusion
In this paper, we proposed a new algorithm named Multi-layer Classification Includ-
ing Verb Consideration (MCIVC) for working on patent technical contradictions clas-
sifying based on Engineering Parameters. This multi-layer classification algorithm
adopts the associated rule-based approach combining the lazy learning. We also de-
signed the algorithm VISAT included in the MCIVC to generate meaningful termsets,
it considers the syntactic structure between words in detail. The MCIVC algorithm
does not only consider the syntactic structure between words, but also consider the
semantic relationship among terms. Classifying patent technical contradictions based
on Engineering Parameters is a useful and new issue. In fact, we have not found any
research classifying patents contradictions based on Engineering Parameters in our
survey. The experimental results showed that the MCIVC can achieve the good
efficacy. Moreover, our algorithm which adopts the VISAT to generate candidate
termsets performs better than comparing algorithms which adopt other methods to
generate candidate termsets. In the future, we will consider whether there are some
relationships among some special Engineering Parameters and try to utilize these
relation-ships to improve the performance of the MCIVC algorithm.
References
1. Altshuller, G.S.: 40 principles: TRIZ keys to tech-nical innovation, translated and edited
by Shulyak, L., Rodman, S., Technical Innovation Center, Worcester, MA (1997)
2. Altshuller, G.: The innovation algorithm: TRIZ, systematic innovation and technical crea-
tivity. Technical Innovation Center (1999)
3. Altshuller, G.S.: Creativity as an Exact Science: The Theory of the Solution of Inventive
Problems. Gordon and Breach Science, New York (1984)
4. Computer aided innovation system: InventionTool 3.0, Chinese patent of software,
2006SR13729 (2006)
5. TechOptimizer (March 2009), http://www.invention-machine.com
6. Pro/Innovator (March 2009), http://www.iwint.com.cn/
7. CREAX INNOVATION SUITE 3.1 (March 2009),
http://www.creaxinnovationsuite.com/
8. Goldfire (March 2009), http://gli.goldfire.com
Search WWH ::




Custom Search