Technology and knowledge document cluster analysis for enterprise R&D strategic planning

Abstract

Through technology and knowledge document analysis, companies can realise the condition of specified technology development and the potential competitors in the market. As patents provide exclusive right and legal protection for patent inventors, patents play an important role in the development of technology. This paper presents the process of patent knowledge extraction and methodologies of patent analysis to improve the efficiency of patent analysis. Furthermore, the methodologies proposed in this paper include patent map analysis, patent technology clustering, patent document clustering and technology maturity measurement. Through these methodologies, companies can gain rich information and perform better in patent management. Moreover, the strategic plans of R&D can also be determined using the methodologies proposed in this paper.

Collective intelligence applied to legal e-discovery: A ten-year case study of Australia franchise and trademark litigation

“…The number of words in a set of documents is used to normalize the value of term frequency. For instance, in patent analysis, text mining techniques including text segmentation, summary extraction, feature selection, term association, cluster generation, topic identification and information mapping are commonly applied [10], [25]. Trappey et al [23] combined the techniques of ontology-based text mining and data mining to identify patent sub-technologies.…”

“…Duplicate cases were removed and the list of all related litigation documents over the recent 10 years are shown in Appendix A. The following steps of the CLI analysis require text mining and a clustering algorithm to identify similar cases within clustered groups and eventually form the clusters with significantly different features [10].…”

“…Given the statistics from each cluster, the key phrase that best describes a given cluster is used to find the associated legal cases and to derive a summary litigation cluster. The litigation summaries are written and based upon the key concepts derived from the key phrases and their relationships clustered using K-Means, a non-hierarchical clustering approach [10,23].…”

A Fuzzy Ontological Knowledge Document Clustering Methodology

“…The Levenshtein distance algorithm and the fuzzy c-mean algorithm were jointly applied to generate the clusters. Another example of clustering for text analysis and synthesis was demonstrated by Hsu et al [14] who used the K-means approach for clustering patent documents.…”

Comparing LDA with pLSI as a Dimensionality Reduction Method in Document Clustering

“…With respect to this type of application, clustering is expected to provide a result at query time. In contrast, enterprise documents stored in the intranet or the patent documents relating to a specific technical field form a document set which is not so small as a search result and, simultaneously, not so large as those targeted by open Web search services [12][15] [18] In this paper, we consider applications managing this type of document set, i.e., a document set of middle-range size and focus on latent Dirichlet allocation (LDA) [10] along with probabilistic latent semantic indexing (pLSI) [17], which are applicable to such document sets in realistic execution time. These two methods share the following special feature: topic multiplicity of each document is explicitly modeled.…”