Paul-Armand Verhaegen

Identifying candidates for design-by-analogy

Although supported by extensive anecdotal evidence, only recently design-by-analogy has been proven to occur often in problem solving and idea generation. However, the circumstances which facilitate problem solving and creative idea generation by analogy are not well understood and most analogies are not developed by applying a formal design-by-analogy methodology. Furthermore, most software tools which aid in finding and/or explaining analogies are based on manually assembled databases, which require a large amount of interactive work to be constructed and maintained. This paper examines the use of automatically distilled product characteristics, called Product Aspects, as a way to automatically and systematically identify candidate products for design-by-analogy. Case studies illustrate this idea generation methodology for three different target products.

Pairwise-adaptive dissimilarity measure for document clustering

This paper introduces a novel pairwise-adaptive dissimilarity measure for large high dimensional document datasets that improves the unsupervised clustering quality and speed compared to the original cosine dissimilarity measure. This measure dynamically selects a number of important features of the compared pair of document vectors. Two approaches for selecting the number of features in the application of the measure are discussed. The proposed feature selection process makes this dissimilarity measure especially applicable in large, high dimensional document collections. Its performance is validated on several test sets originating from standardized datasets. The dissimilarity measure is compared to the well-known cosine dissimilarity measure using the average F-measures of the hierarchical agglomerative clustering result. This new dissimilarity measure results in an improved clustering result obtained with a lower required computational time.

SEABIRD: Scalable search for systematic biologically inspired design

Abstract As more and more people are increasingly turning to nature for design inspiration, tools and methodologies are developed to support the systematic bioideation process. State-of-the-art approaches struggle with expanding their knowledge bases because of interactive work required by humans per biological strategy. As an answer to this persistent challenge, a scalable search for systematic biologically inspired design (SEABIRD) system is proposed. This system leverages experience from the product aspects in design by analogy tool that identifies candidate products for between-domain design by analogy. SEABIRD is based on two conceptual representations, product and organism aspects, extracted from, respectively, a patent and a biological database, that enable leveraging the ever growing body of natural-language biological texts in the systematic bioinspired design process by eliminating interactive work by humans during corpus expansion. SEABIRDs search is illustrated and validated with three well-known biologically inspired design cases.

Topic identification based on document coherence and spectral analysis

Abstract In a world with vast information overload, well-optimized retrieval of relevant information has become increasingly important. Dividing large, multiple topic spanning documents into sets of coherent subdocuments facilitates the information retrieval process. This paper presents a novel technique to automatically subdivide a textual document into consistent components based on a coherence quantification function. This function is based on stem or term chains linking document entities, such as sentences or paragraphs, based on the reoccurrences of stems or terms. Applying this function on a document results in a coherence graph of the document linking its entities. Spectral graph partitioning techniques are used to divide this coherence graph into a number of subdocuments. A novel technique is introduced to obtain the most suitable number of subdocuments. These subdocuments are an aggregation of (not necessarily adjacent) entities. Performance tests are conducted in test environments based on standardized datasets to prove the algorithm’s capabilities. The relevance of these techniques for information retrieval and text mining is discussed.

Webcrawling for a Biological Strategy Corpus to Support Biologically-Inspired Design

In the context of a larger effort to develop a tool that supports ideation in the early stage of Biologically-Inspired Design, this paper describes how the first important research question is tackled: any scalable approach towards such a tool requires a large corpus of biological strategies. This corpus should contain as much of the world’s knowledge about how organisms tackle problems as possible and it should be updated in an automated way. However, currently such a resource or system does not exist. This paper presents a scalable webcrawling approach that allows to continuously search the Internet for biological strategies and to keep its knowledge base up-to-date without manual interaction. The webcrawler solves this needle-in-a-haystack task by combining different classifiers to score the relevance of web documents to the envisaged corpus. It uses these scores to focus future crawling and to gain efficiency. In this way, it becomes possible to continuously harvest new biological strategy documents in a scalable way. Finally, the possible applications of this contribution are positioned in the different existing approaches for systematic BID.

Automatically Populating the Biomimicry Taxonomy for Scalable Systematic Biologically-Inspired Design

Although Biologically-Inspired Design (BID) is gaining popularity, state-of-the-art approaches for systematic BID are still limited by the required interactive work which is proportional to the applied biological database size. This interactive work, depending on the adopted methodology, might encompass model instantiation for each strategy in the biological database, classification into a predefined scheme or extensive result filtering. This contribution presents a first scalable approach to systematic BID with the potential to leverage large numbers of biological strategies. First, a focused webcrawler, based on a combination of Support Vector Machines (SVM), continuously searches for biological strategies on the Internet. The solution to this needle-in-a-haystack task is shown to produce biological strategies interesting for cross-domain Design-by-Analogy (DbA). These resources are then automatically positioned into Ask Nature’s well-known Biomimicry Taxonomy; a 3-level hierarchical classification scheme that enables designers to identify biological strategies relevant to their specific design problem. This paper details the architecture of the proposed system, and presents results indicating the feasibility of the applied approach.© 2012 ASME

A scalable approach for ideation in biologically inspired design

Abstract This paper presents a bioinspiration approach that is able to scalably leverage the ever-growing body of biological information in natural-language format. The ideation tool AskNature, developed by the Biomimicry 3.8 Institute, is expanded with an algorithm for automated classification of biological strategies into the Biomimicry Taxonomy, a three-level, hierarchical information structure that organizes AskNatures database. In this way, the manual work entailed by the classification of biological strategies can be alleviated. Thus, the bottleneck is removed that currently prevents the integration of large numbers of biological strategies. To demonstrate the feasibility of building a scalable bioideation system, this paper presents tests that classify biological strategies from AskNatures reference database for those Biomimicry Taxonomy classes that currently hold sufficient reference documents.

Mention and focus organism detection and their applications for scalable systematic bio-ideation tools

More and more approaches for systematic biologically inspired design (BID) aim to scalably leverage large biological databases. To support the scalable systematic BID process, an automated method for mention and focus organism (FO) detection in biological strategy documents is proposed and validated to perform with 85% precision and 81% recall. Furthermore, a number of potential applications of mention and FO detection are presented, and the biodiversity of two corpora is measured.

Automatically Characterizing Products through Product Aspects

Like most front-end design and problem solving methodologies, TRIZ requires users to abstract their specific system or problem, analyze it through the methodology, and, if applicable, map it back to a specific situation. A methodology and algorithm are proposed that can eliminate this subjective and arduous mapping by formalizing automatically identified, fine-grained product dimensions or Product Aspects. These Product Aspects allow for automatic product characterization, which is a key technology to enable different automated functionalities in idea generation and problem solving contexts, such as automated trend analysis and searching for similar products.

Analysis of Automatic Online Lead User Identification

Lead user identification is a systematic approach to uncovering product development opportunities by identifying lead users, individuals or groups actively involved in modifying or developing products for personal benefit. In this paper, a systematic approach called Fast Lead User IDentification (FLUID) based on online data mining, specifically of the Twitter micro-blogging site, is proposed. Topic classification, sentiment and intent of a given tweet or user-metadata can be automatically determined using various text mining techniques. The described FLUID system makes use of such techniques to rank retrieved users based on indexes derived from well-established lead user characteristics. In the initial analysis phase collection of relevant artifacts and contextual inquiry allow for measuring impact of each index toward delineating lead users from other non-lead users. Through refinement based on statistical analysis of expert assessments the effectiveness of the FLUID system is optimized.