《电子商务 E-business》阅读文献：Mining ideas from textual information

Expert Systems with Applications 37(2010)7182-7188 Contents lists available at Science Direct Expert Systems with Applications ELSEVIER journalhomepagewww.elsevier.com/locate/eswa Mining ideas from textual information Dirk Thorleuchter a, Dirk van den poel b. Anita Prinze b c hofer INT, D-53879 Euskirchen, Appelsgarten 2, Germany ent Universiry, B-9000 Gent, Tweekerkenstraat 2, Belgium Visiting Researcher, Manchester Business School, The University of Manchester, Manchester M15-6PB, Booth Street West, UK ARTICLE INFO A BSTRACT eywords This approach introduces idea mining as process of extracting new and useful ideas from unstructured text. We use an idea definition from technique philosophy and we focus on ideas that can be used to solve Text classification , The rationale for the idea mining approach is taken over from psychology and cognitive science and follows how persons create ideas. To realize the processing, we use methods from text mining and text classification(tokenization, term filtering methods, Euclidean distance measure etc. ) and combine them ith a new heuristic measure for mining ideas As a result, the idea mining approach tomatically new and useful ideas from an user given text. We present these problem solution ing. This approach is evaluated with pat and it is realized as a web-based application, named Technological ldea Miner' that can be used for further testing and evaluation. G 2010 Elsevier Ltd. All rights reserved. 1 Introduction with how persons create ideas especially for problem solution. Therefore, in Section 2 we focus on a general process of creating 11. Overview problem solution ideas and use it as rationale for the idea mining approach. An idea is an image existing or formed in the mind but it can be In recent years, data and text mining techniques explore and written down as textual information. In the last years, we see a analyze huge amounts of available textual data( Coussement continually increasing amount of information. About 80% off all Van den Poel, 2009). Idea mining uses known methods from these this information is stored in textual form(Gentsch Hanlein, techniques and combine them with a new method to create text 999). Examples are research papers, articles in technical periodi- patterns and a new heuristic measure for mining ideas to realize cals, reports, documents, web pages etc. These texts possibly con- the rationale. Therefore, we present the processing of the idea min tain many new ideas. A new idea is often needed to discover ing approach in Section 3 and we introduce this new idea mining unconventional approaches e.g. to create a technological break- measure in Section 4. through. However, a manual extraction of new ideas from these A further task of idea mining is to present the extracted ideas in asses of texts is time consuming and costly Therefore, it is useful a comprehensible way to the user. Therefore, we focus on results of to search for new problem solution ideas automatically comprehensibility research and their relations to our task(see Sec Text mining or knowledge discovery from texts refers generally tion 5). Additionally, we provide an extensive evaluation to show to the process of extracting interesting information and knowledge the success of the idea mining approach and specifically the heuris- from unstructured text(Hotho Nurnberger, Paal, 2005). Refer- tic idea mining measure(see Section 7). ring to this, we introduce idea mining as an automatically process of extracting new and useful ideas from unstructured text using 1. 2. ldea definition Creating ideas is a well-known topic that is related to psychol- ogy and cognitive science. There, we find many approaches dealing two reasons. Firstly, the technological language is much more stan- dardized than the colloquial language(Hoffmann, Kalverkamper Corresponding author at: Fraunhofer INT, Appelsgarten 2, 53879 Euskirchen, Wiegand, 1998: Martin-Bautista, Sanches, Serrano, vila, 2004). 11838305. Therefore, we get better results by analyzing technological texts E-mail address: Dirk. Thorleuchter@int fraunhofer de(d. thorleuchter. with text mining approaches. Secondly our idea definition is taken PhD Candidate, Ghent University. Belgium. over from technique philosophy( rohpohl, 1996). There, an idea is 0957-4174 front matter o 2010 Elsevier Ltd. All rights reserved. oi:10.1016/eswa201004.013

Mining ideas from textual information Dirk Thorleuchter a,*,1, Dirk Van den Poel b , Anita Prinzie b,c a Fraunhofer INT, D-53879 Euskirchen, Appelsgarten 2, Germany bGhent University, B-9000 Gent, Tweekerkenstraat 2, Belgium c Visiting Researcher, Manchester Business School, The University of Manchester, Manchester M15-6PB, Booth Street West, UK article info Keywords: Idea mining Text mining Text classification Technology abstract This approach introduces idea mining as process of extracting new and useful ideas from unstructured text. We use an idea definition from technique philosophy and we focus on ideas that can be used to solve technological problems. The rationale for the idea mining approach is taken over from psychology and cognitive science and follows how persons create ideas. To realize the processing, we use methods from text mining and text classification (tokenization, term filtering methods, Euclidean distance measure etc.) and combine them with a new heuristic measure for mining ideas. As a result, the idea mining approach extracts automatically new and useful ideas from an user given text. We present these problem solution ideas in a comprehensible way to support users in problem solv￾ing. This approach is evaluated with patent data and it is realized as a web-based application, named ‘Technological Idea Miner’ that can be used for further testing and evaluation. 2010 Elsevier Ltd. All rights reserved. 1. Introduction 1.1. Overview An idea is an image existing or formed in the mind but it can be written down as textual information. In the last years, we see a continually increasing amount of information. About 80% off all this information is stored in textual form (Gentsch & Hänlein, 1999). Examples are research papers, articles in technical periodi￾cals, reports, documents, web pages etc. These texts possibly con￾tain many new ideas. A new idea is often needed to discover unconventional approaches e.g. to create a technological break￾through. However, a manual extraction of new ideas from these masses of texts is time consuming and costly. Therefore, it is useful to search for new problem solution ideas automatically. Text mining or knowledge discovery from texts refers generally to the process of extracting interesting information and knowledge from unstructured text (Hotho, Nürnberger, & Paaß, 2005). Refer￾ring to this, we introduce idea mining as an automatically process of extracting new and useful ideas from unstructured text using text mining methods. Creating ideas is a well-known topic that is related to psychol￾ogy and cognitive science. There, we find many approaches dealing with how persons create ideas especially for problem solution. Therefore, in Section 2 we focus on a general process of creating problem solution ideas and use it as rationale for the idea mining approach. In recent years, data and text mining techniques explore and analyze huge amounts of available textual data (Coussement & Van den Poel, 2009). Idea mining uses known methods from these techniques and combine them with a new method to create text patterns and a new heuristic measure for mining ideas to realize the rationale. Therefore, we present the processing of the idea min￾ing approach in Section 3 and we introduce this new idea mining measure in Section 4. A further task of idea mining is to present the extracted ideas in a comprehensible way to the user. Therefore, we focus on results of comprehensibility research and their relations to our task (see Sec￾tion 5). Additionally, we provide an extensive evaluation to show the success of the idea mining approach and specifically the heuris￾tic idea mining measure (see Section 7). 1.2. Idea definition We limit our approach to the technological language because of two reasons. Firstly, the technological language is much more stan￾dardized than the colloquial language (Hoffmann, Kalverkämper, & Wiegand, 1998; Martin-Bautista, Sanches, Serrano, & Vila, 2004). Therefore, we get better results by analyzing technological texts with text mining approaches. Secondly, our idea definition is taken over from technique philosophy (Rohpohl, 1996). There, an idea is 0957-4174/$ - see front matter 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.eswa.2010.04.013 * Corresponding author at: Fraunhofer INT, Appelsgarten 2, 53879 Euskirchen, Germany. Tel.: +49 2251 18305; fax: +49 2251 18 38 305. E-mail address: Dirk.Thorleuchter@int.fraunhofer.de (D. Thorleuchter). 1 PhD Candidate, Ghent University, Belgium. Expert Systems with Applications 37 (2010) 7182–7188 Contents lists available at ScienceDirect Expert Systems with Applications journal homepage: www.elsevier.com/locate/eswa

D. Thorleuchter et aL Expert Systems with Applications 37(2010)7182-7188 defined as a combination of two things: a mean and an appertain- text as described in Section 1. 2. Therefore, with an automatically ing purpose. An example for an idea is a transistor. A transistor is a process, we extract a very large number of overlapping text semiconductor device. It can be used to amplify or switch elec- phrases from the new text. In the remainder of this paper, text tronic signals. Here, we have a mean(a semiconductor device) phrases will be named text patterns In the third step, all extract nd an appertaining purpose (to amplify or switch electronic text patterns are evaluated for novelty and usefulness. This means, signals ). hey are compared to the problem description by using a specific In general, we talk about a new idea if a know mean is related to idea mining measure. With this measure, text patterns can be clas- n unknown purpose or if a known purpose is related to an un- sified as new and useful idea. Therefore, idea mining identifies new known mean(Thorleuchter, Van den poel, Prinzie, 2010c). Then, and useful ideas in three steps: a new idea is a nanomagnet because a nanomagnet is a miniat ized magnet that also can be used to amplify or switch electronic Preparation of a problem description signals. Here we have an unknown mean(a miniaturized magnet) Extraction of text patterns from a new text ppearing together with a known purpose. This new idea could be Evaluation of text patterns for novelty and usefulness concern- useful to humans who are working in the field of electronic signals ing problem description. because in future nanomagnetic technology possibly could replace transistor technology Therefore, we define a new and probably useful idea as a text phrase. This text phrase consists of domain specific terms that oc Fig 1 shows the processing of the idea mining approach in dif- ferent steps based on the rationale for the idea mining process(see into two subsets. The first subset should represent a known mean Section 2). (or a known purpose)and the second subset should represent an With tokenization(Coussement Van den Poel, 2008), texts are unknown purpose(or an unknown mean). Additionally, all terms separated in terms and the term unit is word. The set of different in the first subset should occur together in a text phrase of the terms in a text is reduced by using stop word filtering methods echnological problem description. and stemming( Hotho et al., 2005). For this, a general list of stop words is used as well as the well-known Porter stemming algo rithm( Porter, 1980). 2. Rationale behind idea mining A related problem to the use of stemming is to identify syn- onyms and homonyms. Synonyms are different words with identi- Creating ideas is a well-known topic that is related to creativity cal or at least similar meanings. Homonyms are groups of words n psychology and cognitive science. One of the first descriptions of with the same spelling but with different meanings With stem- the creative process was published by Wallas(1926). His stage ming synonyms and homonyms cannot be identified because model explains creative insights and illuminations for finding a stemming does not use knowledge of the context of a term. In this problem solution. This model consists of a four stages process. In idea mining approach, we do not identify synonyms and hom- age one 'preparation, the problem is analyzed so that a person onyms. This is because the approach always considers the context cognizes the problems dimensions. The stage two 'incubation/ of a term by working on text patterns containing several co-occur- timation and the stage three ' illumination transfer the problem ring terms as described below. from the conscious to the unconscious mind. the unconscious Here, we show how to create these text patterns automatically nind works on the problem continuously and it probably finds a Around each appearance of each term in the new text, we create a solution by creative insights and illuminations. This solution is text pattern containing the selected term and all terms, which oc- transferred to the conscious mind, which means after some time cur in the left and right context of the selected term. To reduce the the person suddenly gets an idea that is new for him and that prob- number of text patterns, we only create text patterns around non ably solves the problem. In the last stage 'verification the idea is sted for novelty and usefulness. New Text Problem description One of the best-known pragmatic approaches of using practical creativity is brainstorming from Osborn(1948). The first step in Tokenization brainstorming is to define the problem e. g. by creating descriptions of the problem. Then, persons generate new ideas using creativity Term fiitering methods like idea association etc. The last step in the brainstorm- ing process is to cluster the generated ideas and to evaluate it for Creating text pattens Creating text patterns novelty and usefulness. Beside this, there are several further approaches dealing with the creation of new ideas. We can learn from all these approaches Creating term vectors Creating term vectors that for creating ideas three steps are necessary. The first step is to ocus on a problem, the second step is to generate some new ideas Euclidean distance measure specific for this problem with creative methods and the third step is to evaluate the generated ideas for novelty and usefulness con- ea mining measure rning quate rationale for the idea mining process. Therefore, idea mining also consists of three steps. In the first step, we focus on the problem. Here, the user of our idea mining approach has to provide textual infor-. mation where he describes his specific problem (a problem and term filtering, text patterns are created and term vectors are built represe description).In the second step, the user has to provide further tex- these text patterns. Term vectors from the new text are compared to term tual information where he supposes the existence of new and use- vectors from the ne om the problem description using the Euclidean distance measure. Then, term re compared to their most similar term vectors from the ful ideas(a new text)that probably can solve his problem( Ripke problem description using the idea mining measure. As a Stober, 1972). Ideas are contained in text phrases inside this new from the new text that represent new and useful ideas esult, we get term vectors

defined as a combination of two things: a mean and an appertain￾ing purpose. An example for an idea is a transistor. A transistor is a semiconductor device. It can be used to amplify or switch elec￾tronic signals. Here, we have a mean (a semiconductor device) and an appertaining purpose (to amplify or switch electronic signals). In general, we talk about a new idea if a know mean is related to an unknown purpose or if a known purpose is related to an un￾known mean (Thorleuchter, Van den Poel, & Prinzie, 2010c). Then, a new idea is a nanomagnet because a nanomagnet is a miniatur￾ized magnet that also can be used to amplify or switch electronic signals. Here we have an unknown mean (a miniaturized magnet) appearing together with a known purpose. This new idea could be useful to humans who are working in the field of electronic signals because in future nanomagnetic technology possibly could replace transistor technology. Therefore, we define a new and probably useful idea as a text phrase. This text phrase consists of domain specific terms that oc￾cur together in textual information. These terms can be divided up into two subsets. The first subset should represent a known mean (or a known purpose) and the second subset should represent an unknown purpose (or an unknown mean). Additionally, all terms in the first subset should occur together in a text phrase of the technological problem description. 2. Rationale behind idea mining Creating ideas is a well-known topic that is related to creativity in psychology and cognitive science. One of the first descriptions of the creative process was published by Wallas (1926). His stage model explains creative insights and illuminations for finding a problem solution. This model consists of a four stages process. In stage one ‘preparation’, the problem is analyzed so that a person recognizes the problem’s dimensions. The stage two ‘incubation/ intimation’ and the stage three ‘illumination’ transfer the problem from the conscious to the unconscious mind. The unconscious mind works on the problem continuously and it probably finds a solution by creative insights and illuminations. This solution is transferred to the conscious mind, which means after some time the person suddenly gets an idea that is new for him and that prob￾ably solves the problem. In the last stage ‘verification’, the idea is tested for novelty and usefulness. One of the best-known pragmatic approaches of using practical creativity is brainstorming from Osborn (1948). The first step in brainstorming is to define the problem e.g. by creating descriptions of the problem. Then, persons generate new ideas using creativity methods like idea association etc. The last step in the brainstorm￾ing process is to cluster the generated ideas and to evaluate it for novelty and usefulness. Beside this, there are several further approaches dealing with the creation of new ideas. We can learn from all these approaches that for creating ideas three steps are necessary. The first step is to focus on a problem, the second step is to generate some new ideas specific for this problem with creative methods and the third step is to evaluate the generated ideas for novelty and usefulness con￾cerning the problem. Referring to these approaches, we build an adequate rationale for the idea mining process. Therefore, idea mining also consists of three steps. In the first step, we focus on the problem. Here, the user of our idea mining approach has to provide textual infor￾mation where he describes his specific problem (a problem description). In the second step, the user has to provide further tex￾tual information where he supposes the existence of new and use￾ful ideas (a new text) that probably can solve his problem (Ripke & Stöber, 1972). Ideas are contained in text phrases inside this new text as described in Section 1.2. Therefore, with an automatically process, we extract a very large number of overlapping text phrases from the new text. In the remainder of this paper, text phrases will be named text patterns. In the third step, all extracted text patterns are evaluated for novelty and usefulness. This means, they are compared to the problem description by using a specific idea mining measure. With this measure, text patterns can be clas￾sified as new and useful idea. Therefore, idea mining identifies new and useful ideas in three steps: Preparation of a problem description Extraction of text patterns from a new text and Evaluation of text patterns for novelty and usefulness concern￾ing problem description. 3. Idea mining process Fig. 1 shows the processing of the idea mining approach in dif￾ferent steps based on the rationale for the idea mining process (see Section 2). With tokenization (Coussement & Van den Poel, 2008), texts are separated in terms and the term unit is word. The set of different terms in a text is reduced by using stop word filtering methods and stemming (Hotho et al., 2005). For this, a general list of stop words is used as well as the well-known Porter stemming algo￾rithm (Porter, 1980). A related problem to the use of stemming is to identify syn￾onyms and homonyms. Synonyms are different words with identi￾cal or at least similar meanings. Homonyms are groups of words with the same spelling but with different meanings. With stem￾ming synonyms and homonyms cannot be identified because stemming does not use knowledge of the context of a term. In this idea mining approach, we do not identify synonyms and hom￾onyms. This is because the approach always considers the context of a term by working on text patterns containing several co-occur￾ring terms as described below. Here, we show how to create these text patterns automatically. Around each appearance of each term in the new text, we create a text pattern containing the selected term and all terms, which oc￾cur in the left and right context of the selected term. To reduce the number of text patterns, we only create text patterns around non￾Fig. 1. Processing of our idea mining approach in different steps: After tokenization and term filtering, text patterns are created and term vectors are built representing these text patterns. Term vectors from the new text are compared to term vectors from the problem description using the Euclidean distance measure. Then, term vectors from the new text are compared to their most similar term vectors from the problem description using the idea mining measure. As a result, we get term vectors from the new text that represent new and useful ideas. D. Thorleuchter et al. / Expert Systems with Applications 37 (2010) 7182–7188 7183

D. Thorleuchter et al. Expert Systems with Applications 37(2010)7182-7188 stop words and around terms that occur both in the new text and in the problem description. (w+k)≥1)v(+j=n)hi∈{1…,n,(2) One important decision to be taken is to determine the length of a text pattern. Text patterns should not be too small so that they tef=m>(w)≥)v-j=1)m∈1…,n(3 contain all terms representing a new idea. Further text patterns that are related to the new idea. For example if we set the length of After computing Pe and rg r, we can build a text pattern T, around the text patterns to l then a text pattern contains the selected term, the term w from the text t=[w1,.,Wnl I terms from its left context and also I terms from its right context. The cardinality of the set of stop word filtered and stemmed terms Ti=w -tort,., Wi,..., wi+1 from this pattern is normally smaller than 2*1+1 because some terms are stop words, some terms occur twice and some terms For each text pattern from the new text, we create a term vector in have the same stem vector space model. The size of the vector is defined by the number In this paper, we do not use a constant length I for all patterns of different stemmed and stop word filtered terms in the new text. but a variable length of text patterns based on a dynamic adapta- for text pattern encoding, we use binary term vectors that means a tion of its context. This is realized by using a term vector element is set to one if the corresponding unstemmed term heme based on the difference between stop words and non-stop is used in the text pattern and to zero if the term is not. We also ce of a stop w not as high as the importance of a non-stop word If an author for build text patterns from the problem description and create term mulates an idea very briefly by joining catchwords together then vectors as described above. he normally does not use many stop words and the text pattern ing measure. This idea mining measure is described in Section 4. By style that means his writing is not expressed in a clear and simple description, we can compute a result value always between 0% way then he normally uses more stop words and the text pattern and 100% using this measure. The greater the result value the more ngth has to be larger. In the idea mining application the value is the probability that the vector from the new text represents a of text pattern length I and the percentage of the importance of stop words u and of non-stop words v can be provided by the user. new and useful idea concerning a vector from the problem To compute the variable length of a text pattern, we firstly de- fine the term weighting scheme Ne use this measure for comparing vectors from the new text to their most similar vectors from the problem description but Definition 1. Let (a text)T=wi,, Wan]be a list of terms(words) vector to its most similar vectors predominate result values from w, in order of appearance and let n e N be the number of terms int comparing a vector to its further vectors. For example if a vector stop terms(Thorleuchter, Van den Poel, Prinzie, 2010b)and let from the new text is similar to one from the problem descripe y then the idea is not new to the user regardless whether result m E N be the number of terms in 2. Let the percentage ueN be a ues from comparing this vector to further vectors from the prob- term weighting coefficient for stop words. Let the percentage ve lem description are greater than zero. therefore, we can be sure be a term weighting coefficient for non-stop words. Then, we define f(wi)EN as term weighting scheme that a vector represents a new and useful idea only if it gets a great result value from idea mining measure concerning one of its most fs( wi) uw∈E wife(i∈ (1) similar vectors. Further, the computing of the idea mining measure is time consuming. Therefore, it is necessary to limit the number of comparisons with idea mining measure for implementing an idea We give an example for this. The text pattern components for mining application. equency conversion of infrared lasers' is built around the word We choose a two-step classification way. In the first step. conversion. It contains the word conversion itself, three terms compare each vector from the new text to all vectors from the from its left context(components for frequency ), and three terms problem description by using the well-known Euclidean distance from its right context (of infrared lasers). Here, we use a constant measure. Fortunately, the computing of the Euclidean distance ngth 1-3 and a term weighting scheme with a-B=100%. This measure is not time consuming so that it is suited for implement means the importance of a stop word is equal to the importance ing in an idea mining application In detail, for each vector from the of a non-stop word. The next text pattern is an example for a var- new text, we identify all vectors from the problem description iable length: 'ln a lst phase, known but so far not available mate where the euclidean distance result value is the lowest that means rials and technologies such as layer systems and crystals. This text we identify the most similar vectors In the second step we com- pattern is built around the word'technologies Here we use a con- pare each vector from the new text to its most similar vectors using stant length 1=3 and a term weighting scheme with u= 10% and the idea mining measure 100%. As a result, this text pattern contains six terms from Each vector from the new text that is compared to several the right context and eleven terms from the left context of the term similar vectors -gets the highest result value from idea mining technologies. In this example, non-stop words are phase, materi- measure as result value. To identify a new and useful idea we als, technologies, layer, systems, and crystal. We compute the use alpha-cut method An alpha-cut of the idea mining measure re- number of terms from the left and right context as described sult value is the set of all vectors from the new text such that the appertaining result value is greater than or equal to alpha(a). In the idea mining application, the user can provide the value of Definition 2. Letle N be a constant length of text patterns. Let et be the number of terms from the left context of a text pattern that 4 ldea mining measure built around the term w. Let hgi be the number of terms from he right context of a text pattern that is built around the term w With the idea mining measure, we compare a vector that repre- Then, we define t E N and Aght EN as ents a text pattern from the new text to its most similar vectors

stop words and around terms that occur both in the new text and in the problem description. One important decision to be taken is to determine the length of a text pattern. Text patterns should not be too small so that they contain all terms representing a new idea. Further text patterns should not be too large so that only terms occur in the text patterns that are related to the new idea. For example if we set the length of the text patterns to l then a text pattern contains the selected term, l terms from its left context and also l terms from its right context. The cardinality of the set of stop word filtered and stemmed terms from this pattern is normally smaller than 2 l + 1 because some terms are stop words, some terms occur twice and some terms have the same stem. In this paper, we do not use a constant length l for all patterns but a variable length of text patterns based on a dynamic adapta￾tion of its context. This is realized by using a term weighting scheme based on the difference between stop words and non-stop words because the importance of a stop word in a text pattern is not as high as the importance of a non-stop word. If an author for￾mulates an idea very briefly by joining catchwords together then he normally does not use many stop words and the text pattern length can be small. If an author formulates an idea in a flowery style that means his writing is not expressed in a clear and simple way then he normally uses more stop words and the text pattern length has to be larger. In the idea mining application the value of text pattern length l and the percentage of the importance of stop words u and of non-stop words v can be provided by the user. To compute the variable length of a text pattern, we firstly de- fine the term weighting scheme. Definition 1. Let (a text) T ¼ ½w1; ... ; wn be a list of terms (words) wi in order of appearance and let n 2 N be the number of terms in T and i 2 ½1; ... ; n. Let R ¼ fw~ 1; ... ; w~ mg be a set of domain specific stop terms (Thorleuchter, Van den Poel, & Prinzie, 2010b) and let m 2 N be the number of terms in R. Let the percentage u 2 N be a term weighting coefficient for stop words. Let the percentage v 2 N be a term weighting coefficient for non-stop words. Then, we define fgðwiÞ 2 N as term weighting scheme: fg ðwiÞ ¼ ujwi 2 R vjwi R R ð8i 2 f1; ... ; ngÞ: ð1Þ We give an example for this. The text pattern ‘components for frequency conversion of infrared lasers’ is built around the word ‘conversion’. It contains the word conversion itself, three terms from its left context (components for frequency), and three terms from its right context (of infrared lasers). Here, we use a constant length l = 3 and a term weighting scheme with a = b = 100%. This means the importance of a stop word is equal to the importance of a non-stop word. The next text pattern is an example for a var￾iable length: ‘In a 1st phase, known but so far not available mate￾rials and technologies such as layer systems and crystals’. This text pattern is built around the word ‘technologies’. Here we use a con￾stant length l = 3 and a term weighting scheme with u = 10% and v = 100%. As a result, this text pattern contains six terms from the right context and eleven terms from the left context of the term ‘technologies’. In this example, non-stop words are phase, materi￾als, technologies, layer, systems, and crystal. We compute the number of terms from the left and right context as described below: Definition 2. Let l 2 N be a constant length of text patterns. Let l left i be the number of terms from the left context of a text pattern that is built around the term wi. Let l right i be the number of terms from the right context of a text pattern that is built around the term wi. Then, we define l left i 2 N and l right i 2 N as: l right i ¼ min j X j k¼1 fg ðwiþkÞ P l ! _ ði þ j ¼ nÞ 8i 2 f g 1; ... ; n ; ð2Þ l left i ¼ min j X j k¼1 fg ðwikÞ P l ! _ ði  j ¼ 1Þ 8i 2 f g 1; ... ; n : ð3Þ After computing l left i and l right i , we can build a text pattern Ti around the term wi from the text T ¼ ½w1; ... ; wn. Ti ¼ wil left i ; ... ; wi; ... ; wiþl right i h i: ð4Þ For each text pattern from the new text, we create a term vector in vector space model. The size of the vector is defined by the number of different stemmed and stop word filtered terms in the new text. For text pattern encoding, we use binary term vectors that means a vector element is set to one if the corresponding unstemmed term is used in the text pattern and to zero if the term is not. We also build text patterns from the problem description and create term vectors as described above. To identify new and useful ideas, we create a specific idea min￾ing measure. This idea mining measure is described in Section 4. By comparing a vector from the new text to one from the problem description, we can compute a result value always between 0% and 100% using this measure. The greater the result value the more is the probability that the vector from the new text represents a new and useful idea concerning a vector from the problem description. We use this measure for comparing vectors from the new text to their most similar vectors from the problem description but not to all vectors. This is because result values from comparing a vector to its most similar vectors predominate result values from comparing a vector to its further vectors. For example if a vector from the new text is similar to one from the problem description then the idea is not new to the user regardless whether result val￾ues from comparing this vector to further vectors from the prob￾lem description are greater than zero. Therefore, we can be sure that a vector represents a new and useful idea only if it gets a great result value from idea mining measure concerning one of its most similar vectors. Further, the computing of the idea mining measure is time consuming. Therefore, it is necessary to limit the number of comparisons with idea mining measure for implementing an idea mining application. We choose a two-step classification way. In the first step, we compare each vector from the new text to all vectors from the problem description by using the well-known Euclidean distance measure. Fortunately, the computing of the Euclidean distance measure is not time consuming so that it is suited for implement￾ing in an idea mining application. In detail, for each vector from the new text, we identify all vectors from the problem description where the Euclidean distance result value is the lowest that means we identify the most similar vectors. In the second step, we com￾pare each vector from the new text to its most similar vectors using the idea mining measure. Each vector from the new text – that is compared to several similar vectors – gets the highest result value from idea mining measure as result value. To identify a new and useful idea we use alpha-cut method. An alpha-cut of the idea mining measure re￾sult value is the set of all vectors from the new text such that the appertaining result value is greater than or equal to alpha ða~Þ. In the idea mining application, the user can provide the value of a~. 4. Idea mining measure With the idea mining measure, we compare a vector that repre￾sents a text pattern from the new text to its most similar vectors 7184 D. Thorleuchter et al. / Expert Systems with Applications 37 (2010) 7182–7188

D. Thorleuchter et aL Expert Systems with Applications 37(2010)7182-7188 85 from the problem description to identify a new and useful idea in- of unknown terms which occur frequently in the new text. We text pattern from the new text where all terms representing a nhe m3 as measure for frequently occurrence of unknown terms side the text pattern from the new text. In detail, we have to find do new text mean(purpose)and no terms representing a purpose(mean) occur in a text pattern from the problem description. If all terms in the text pattern from the new text are know the idea is not useful if all terms in the text patter ern from the a changing purpose t are unknown because there is no relation to the problem. It is for new ideas shown in Thorleuchter (2008) that to find new and useful ideas the number of known terms(e.g. representing a mean)and the num Definition 6. Let i be a set of these characteristic terms(stemmed er of unknown terms (e.g. representing an appertaining purpose) and stop word filtered). Let 0 E(0.1)be a term vector in vector space model concerning 2. Let t=oi ni=ka10ik. Ok be the number of these characteristic terms in text pattern with number i. Definition 3. Let a, be a the set of stemmed and stop word filtered We define ma as measure for changing means and purposes terms representing a text pattern with number i from the new text. representing a text pattern with number j from the problem ma=I(t>0) Let B be a set of stemmed and stop word filtered term (t=0) description. Let y be the set of all stemmed and stop word filtered The idea mining measure bases on all four heuristic sub terms from the new text. Let x= lyl be the cardinality of ,. Let oiE 10, 1 be a term vector in vector space model concerning a Let pi e 10. 1 be a term vector in vector space model concerning 序Letp=|=∑k=1 Dik be the number of all( known and Definition7.Leth∈{1,……,4} and let gn≥ 0 be weighting factors unknown)terms in text pattern with number i. Let with Zh=1gn=1. Let the idea mining measure be the sum of all q=linbI= Eka1Oik Pik be the number of known terms in text four sub measures multiplied by weighting factors gh in case of pattern with number i concerning a text pattern with numberj P* q from the problem description. Then, we define mI as measure for g1m1 +g2m2+g3m3+ g4m rel-balanced known and unknown term distribution 2(p-q =(q<别 5. Idea mining and comprehensibility research The known terms in the text pattern from the new text should occur in the problem description more frequently than other The aim of idea mining is to find new and useful ideas but also terms. This is because they represent a known mean or a known to present these ideas in a comprehensible way to the user To real purpose that is a central part of the problem. In the problem ize this, we focus on comprehensibility research. description, terms that represent the problem occur more fre- Up to the 1960s comprehensibility was a property of the text. It quently than other terms. For this, we define these frequent terms was measured in an objective way by analyzing text parameters by using a percentage z as parameter and we compute m2 as the like word length, sentence length, word-usability, relationship be- number of known and frequent terms over the number of all tween number of different words and number of words. The well- known terms known approach in this time was the 'Reading Ease -formula from Flesch(1948). Later research in this field focuses on cognitive effects by doing Definition 4. Let z be a percentage. Let 8 be a set of zo most textual production and reception. The results of this research are frequently stemmed and stop word filtered terms in the probler presented by two approaches: the 'Hamburger Vers- description. Let &E 10, 1)" be a term vector in vector space model tandlichkeitsmodell'(Langer, Schulz v. Thun, Tausch, 1974)and concerning.Letr=|∩房∩l=∑k= Oi k' Pik· k be the num ber of known terms which occur frequently in the probler the 'Groebener Modell(Groeben, 1982). Both approaches describe description We define mz as measure for frequently occurrence four dimensions of comprehensibility: simplicity, structure-org tion, brevity-shortness and interest-liveliness of known terms in the problem description. A further approach from cognition research is named text The unknown terms in the text pattern from the new text excerption. if a human expert finds new and useful ideas in texts he highlights all corresponding text phrases e.g. with text marking. This behaviour is described by Puppe, Stoyan, and Studer (2003). resent a new approach(an unknown mean or purpose), which is a In the idea mining application, text excerption is used to pres- entral part of the new idea. These terms normally occur more fre nt the extracted ideas to the user(Fig. 2 shows an example). For quently than other terms in the new text because this text deals about the new idea. For this, we also define these frequent terms In a second phase, technologies will be selected from them and by using a percentage z as parameter and we compute m3 as the optical nonlinear components meeting specified requirements number of unknown and frequent terms over the number of all un- will be realized in experimental models and tested for durability and suitability as oPo (optical parametric oscillator) or OPA (optic of the project is to demonstrate the feasibility and producibility stoi witon 5 ered t m a se t he zew test freq ento ste ee and o o sc ptd abonoeli onlinear components for infrared ranges from ector in vector space model concerning o. Let S=∩∩=∑x0,飞-∑=1Ok·Pk: be the number ig2 e present the new text back to the user with text pattems in bold print

from the problem description to identify a new and useful idea in￾side the text pattern from the new text. In detail, we have to find text pattern from the new text where all terms representing a mean (purpose) and no terms representing a purpose (mean) occur in a text pattern from the problem description. If all terms in the text pattern from the new text are known, which means all terms also occur in a text pattern from the prob￾lem description then the idea is not new to the user. Furthermore, the idea is not useful if all terms in the text pattern from the new text are unknown because there is no relation to the problem. It is shown in Thorleuchter (2008) that to find new and useful ideas the number of known terms (e.g. representing a mean) and the num￾ber of unknown terms (e.g. representing an appertaining purpose) shall be well balanced. Definition 3. Let ai be a the set of stemmed and stop word filtered terms representing a text pattern with number i from the new text. Let bj be a set of stemmed and stop word filtered terms representing a text pattern with number j from the problem description. Let c be the set of all stemmed and stop word filtered terms from the new text. Let x ¼ jcj be the cardinality of c. Let xi 2 f0; 1gx be a term vector in vector space model concerning ai. Let qj 2 f0; 1gx be a term vector in vector space model concerning bj. Let p ¼ jaij ¼ Px k¼1xi;k be the number of all (known and unknown) terms in text pattern with number i. Let q ¼ jai \ bjj ¼ Px k¼1xi;k  qj;k be the number of known terms in text pattern with number i concerning a text pattern with number j from the problem description. Then, we define m1 as measure for well-balanced known and unknown term distribution. m1 ¼ 2ðpqÞ p q P p 2  ; 2q p q 0Þ; 0 ðt ¼ 0Þ: ð8Þ The idea mining measure bases on all four heuristic sub measures. Definition 7. Let h 2 f1; ... ; 4g and let gh P 0 be weighting factors with P4 h¼1gh ¼ 1. Let the idea mining measure be the sum of all four sub measures multiplied by weighting factors gh in case of p – q. m ¼ g1m1 þ g2m2 þ g3m3 þ g4m4 ðp – qÞ; 0 ðp ¼ qÞ: ð9Þ 5. Idea mining and comprehensibility research The aim of idea mining is to find new and useful ideas but also to present these ideas in a comprehensible way to the user. To real￾ize this, we focus on comprehensibility research. Up to the 1960’s comprehensibility was a property of the text. It was measured in an objective way by analyzing text parameters like word length, sentence length, word-usability, relationship be￾tween number of different words and number of words. The well￾known approach in this time was the ‘Reading Ease’-formula from Flesch (1948). Later research in this field focuses on cognitive effects by doing textual production and reception. The results of this research are presented by two approaches: the ‘Hamburger Vers￾tändlichkeitsmodell’ (Langer, Schulz v. Thun, & Tausch, 1974) and the ‘Groebener Modell’ (Groeben, 1982). Both approaches describe four dimensions of comprehensibility: simplicity, structure-orga￾nization, brevity-shortness and interest-liveliness. A further approach from cognition research is named text excerption. If a human expert finds new and useful ideas in texts he highlights all corresponding text phrases e.g. with text marking. This behaviour is described by Puppe, Stoyan, and Studer (2003). In the idea mining application, text excerption is used to pres￾ent the extracted ideas to the user (Fig. 2 shows an example). For Fig. 2. We present the new text back to the user with text patterns in bold print that represent new and useful ideas. D. Thorleuchter et al. / Expert Systems with Applications 37 (2010) 7182–7188 7185

idea mining approach is evaluated by using ouridea mining ed application and The idea mining measure as central point in th ically founded. Therefore, it is crucial to evaluation to show their baseline because we are not aware of ot gth of the text patterns. The (Ferber, 2003) as ure for the baseli term weighting between stop words and

the ‘Groebener Modell’ marking text pattern is important for struc￾ture-organization and this leads directly to comprehensibility. In this point there are differences between the ‘Groebener Modell’ and the ‘Hamburger Verständlichkeitsmodell’ in which structure￾organization is not so important for comprehensibility. As a result, the presentation of ideas in the idea mining applica￾tion based on text excerption. It is comprehensible after the ‘Groebener Modell’ and it is less comprehensible after the ‘Ham￾burger Verständlichkeitsmodell’. 6. Results and discussions In a study for the German Ministry of Defence (MoD), we use this approach to identify new technological ideas for the German defence research program. In detail, we have to identify new solu￾tion ideas to solve current problems in German defence based re￾search projects. We extract new ideas from 300 descriptions of research projects granted in 2006 by the National Institute of Stan￾dards and Technology (NIST) in the United States Small Business Innovation Research (SBIR) Program. We use textual information from current defence based research projects of the German MoD as problem description (Thorleuchter, Van den Poel, & Prinzie, 2010a). As a result, we extract several new ideas that are useful for German defence research planners and that now are used as starting point for collaboration projects or for new defence based research projects. A proper selection of these ideas is a strategic is￾sue and - together with the weapon selection problem (Dagdevi￾ren, Yavuz, & Kilinc, 2009) – it has significant impacts to the efficiency of future defence systems. The results are published in Fenner and Thorleuchter (2009). Here, we show some successful examples: A modified focal plane array technology is identified that can be used to create a detector for the far ultraviolet spectrum. It leads to an improvement of military reconnaissance. This idea is new be￾cause up to now focal plane array technology is only used in the infrared, visual and near ultraviolet area. Further, the approach identifies personnel ultrasonic locating equipment that was originally developed to make orientation pos￾sible for fire fighters in dense smoke. It also can be used to improve the location and navigation of soldiers in urban warfare (e.g. in buildings). Additionally, the approach shows that the use of avalanche pho￾todiode (APD) technology can improve the internal gain and the dark current of infrared detectors. This also leads to an improve￾ment of military reconnaissance. This study shows that some of the automatically extracted ideas are useful for technological research planners from the German MoD. Unfortunately, the used problem description (textual infor￾mation about current defence based research projects) is classified as German restricted (Verschlusssache – Nur für den Dienstgebr￾auch) that means it is not allowed to distribute it to the scientific community. Therefore, we cannot use the results of this study to evaluate this idea mining approach. However, a separate evalua￾tion (see Section 7) is done using (unclassified) patent data that al￾lows re-computing of the evaluation. 7. Evaluation The idea mining measure as central point in the idea mining ap￾proach consists of four heuristic sub measures that are not theoret￾ically founded. Therefore, it is crucial to provide an extensive evaluation to show their success. We compare this approach to a baseline because we are not aware of other approaches for idea mining. As measure for the baseline, we use Jaccard’s coefficient (Ferber, 2003) as well-known heuristic similarity measure. The idea mining approach is evaluated by using our idea mining application (see Section 8). There the web-based application and all texts that are used for evaluation are presented. Additionally, we create an alternative idea mining application, based on Jac￾card’s coefficient instead of the idea mining measure for the sole purpose of comparison to the baseline. For evaluation, we use patent data because in patent descrip￾tions, we normally can find new ideas, which include a consider￾able part of scientific and technological knowledge (Li, Wang, & Hong, 2009). We use the abstract of a patent as new text. A patent often bases on further patents. We aggregate abstracts of theses references as problem description. Then we identify new and use￾ful ideas from this patent concerning its patent references using the idea mining applications. We use abstracts from 40 randomly selected patents and from their references, a general stop word list and Porter stemmer for evaluation. Then we determine the parameters of the idea mining measure (g1; g2; g3; g4; a~, and z) as well as the parameters for the length of the text patterns (l, u, and v). For this, we use further patent data and their references as new text and as problem description. The results are evaluated by a hu￾man expert and compared to each single sub measure m1, m2, m3 and m4 alone. We find out that using the first sub measure alone is successful. If this sub measure is small then the corresponding text pattern normally does not contain a new and useful idea. If this sub measure is large then the probability that the text pattern con￾tains a new idea is also high. We also find out that using the further sub measures alone is not successful. This means, they are successful only if the result value of the first sub measure is medium to high. Therefore, they only can be used in addition to the first sub measure. The results of the second and third sub measures depend on the parameter z. This parameter is used to define frequent terms by building a set of z% most frequently stemmed and stop word fil￾tered terms. We heuristically think that this parameter should be between 10% and 30% to get good sub measures. This is because if z is greater than 30% then we probably classify several terms, which only occur once as frequent terms. If z is smaller than 10% then we only identify high frequently terms for the set. In this case, the result values of the second and third sub measures are small regardless weather known terms occur frequently in the problem description or unknown terms occur frequently in the new text. Therefore, we determine z to the mean value (20%). Additionally, we see that the second and third sub measure is nearly equally successful and that the fourth sub measure is less successful. Therefore, we heuristically determine the parameters of g1 to 50%, g2 to 20%, g3 to 20% and g4 to 10%. We also have used other values to optimize the combination of these four sub measures. However, we do not find a combination that is generally superior to the selected combination. This is be￾cause the success of these value combinations depends on the quality of the user given textual information. Then, we determine the alpha-cut value a~ of the idea mining measure m. If the percentage a~ is small then we get many result items. This leads to a small precision value because many extracted text patterns do not contain a new and useful idea. If a~ is large then we only get a very small number of results and probably our recall value is small because we do not find most of the new and useful ideas in the new text. A human expert checks the results of several patent descriptions for an optimal value of a~. He gets the experi￾ence that 60% is a good compromise. Therefore, we set a~ to 60%. We also determine the alpha-cut value of Jaccard’s coefficient as measure for the baseline to 20% by using the same way of evalua￾tion as described above. After this, we determine the length of the text patterns. The length depends on the parameter l and on fg(wi), a term weighting scheme that is based on the difference between stop words and 7186 D. Thorleuchter et al. / Expert Systems with Applications 37 (2010) 7182–7188

D. Thorleuchter et aL Expert Systems with Applications 37(2010)7182-7188 non-stop words(see Section 3). Text patterns should not be too These files can be formatted in various ways e.g. as plain text, html, small so that they contain all terms representing a new and useful xml etc. However, scripting code. (html-or xml-)tags, and images ea. Additionally, text patterns should not be too large so that fur- are discarded that means the application extracts plain text from ther terms occur in the text patterns that are not related to the new the provided files. Then, the user has to select the language of these nd useful idea To find out an optimal size of text patterns, we cre- texts to integrate a general stop word list of this language. The ate text patterns from several patent descriptions by using differ- application offers general stop word lists in English, German, ent values for I and for the percentages u and 1. A human expert Dutch, Spain and French. After determining the parameters of the checks the different length of these text patterns for an optimal application the automatically extraction of new and useful ideas size. He gets the best results by setting the value of text pattern from the new text starts as described in the idea mining process length I to 7 terms and the percentage u to 50% and v to 100%. (see Sections 3 and 4). As a result, new ideas are presented as de- Then, the approach extracts automatically about 200 new ideas scribed in Section 5 from the 40 randomly selected patents. To cluster these results, means and purposes are assigned to scientific categories in the sci- 9. Conclusions and future research ence citation index and examples are presented below. Several ideas are identified that uses methods from 'Artificial Intelligence'(mean This study shows the success of an automatic approach for find for applications in'Health Care Sciences and Services'(purpose). We ing new ideas from textual information. For this, the study trans- also identify new ideas using'lmaging Science and Photographic forms creativity approaches from psychology and cognitive ideas use techniques from ' Remote Sensing (mean)in the field of redefine an abstract term (an idea)in a concrete way that it can Tropical Medicine(purpose). Additionally, several ideas use be used for computing with text mining methods In detail, it is Computer Science, Theory and Methods'(mean) for applications nown that a technological idea represents a combination of a pur- in 'Psychiatry'(purpose). Furthermore, methods from 'Artificial Intelligence (mean)are used for Automation and Control Systems' Pose and a mean and that purposes and means are defined by a Additionally, it is shown that problems and problem solution To evaluate these results, we use precision and recall measures ideas can be represented as term vectors in vector space model commonly used in information retrieval based on true positives, For this, the study contributes a new(idea mining)measure.This false positives and false negatives. For this, we have to define the measure identifies new ideas by comparing vectors that represent ground truth for our evaluation. Therefore, a human expert also a problem to vectors that represent a problem solution idea.Last, it identifes new and useful ideas from these patents manually that is shown that approaches from comprehensibility research can be means without using our idea mining approach. He uses the idea adopted to this approach to present the new ideas in a comprehen definition in Section 1.2. This means, he checks each text pattern sible way to the user. As further main finding. it is demonstrated for finding terms representing a known mean(purpose)and terms that this theoretical approach can be realized by a web-based representing an unknown purpose( mean). These results are the application. The success of the idea mining measure is proved by ground truth for the evaluation. omparing it to further heuristic measures(overlap-index For each patent, we compute its precisi d recall values by sine-similarity and dice-similarity) using the idea mining measure and by using the Jaccards coeffi Directions for future research are given by the fact that nowa cient. Then, we compute the average precision and recall val lays there a result, we get a precision value of 40% and a recall value of the internet and this information probably contains many new 25% by using the idea mining approach with the idea mining mea- technological ideas. Enlarging this approach to a web idea mining sure.A precision value of 40% means that if the idea mining ap- approach that automatically identifies problem solution ideas from extracts ten text patterns then four of them represent the internet is an interesting topic for further research d useful idea. a recall value of 25% means that if there four new and useful ideas in the new text then the idea mining ap- and the idea mining measure can probably be enlarged with fur- proach extracts only one of them. In contrast to this, we get a pre- ther aspects to improve its quality that means to get better results cision value of 30% and a recall value of 20% by using Jaccard's for the precision and recall values coefficient. This is because in some texts Jaccard's coefficient ex- A further aspect is to transform this idea mining approach to the tracts text patterns from the new text that are similar to text pat- colloquial language. For this, it is necessary that the idea definition rns from the problem description. This represents probably a also contains new product ideas from the consumers.Then,new known idea but not a new idea product ideas can be identified to support marketing activities. Beside jaccard, s coefficient, we also test other well-known heu- stic measures like overlap-index, cosine-similarity and dice-si Last, the approach can be extended with innovation-related as- pects. Then, extracted ideas can be classified as innovative ideas ilarity(Ferber, 2003)as baseline. However, we get nearly the same and might be used as starting point for the new product results for the precision(30%)and for the recall (20%)value 8. The idea mining application Acknowledgements We thank Joachim Schulze and Jorg Fenner for constructive The idea mining application focus on users without extensive technical comments perts. We give them the possibility to extract specifically problem references They can access to the web-based application via the internet. It is Coussement, K. Van den Poel, D (2008) Integrating availableunderhttp://www.text-mining.infoanditisprogrammed rough call center emails into a decision support system for churn prediction. An user has to provide two textual files, a problem description Coussement, K,& Van den poel, D (200 by integrating emotions from cl and a new text that probably consists of problem solution ideas. multiple classifiers Expert Systems with Applications, 36, 6127-6134

non-stop words (see Section 3). Text patterns should not be too small so that they contain all terms representing a new and useful idea. Additionally, text patterns should not be too large so that fur￾ther terms occur in the text patterns that are not related to the new and useful idea. To find out an optimal size of text patterns, we cre￾ate text patterns from several patent descriptions by using differ￾ent values for l and for the percentages u and v. A human expert checks the different length of these text patterns for an optimal size. He gets the best results by setting the value of text pattern length l to 7 terms and the percentage u to 50% and v to 100%. Then, the approach extracts automatically about 200 new ideas from the 40 randomly selected patents. To cluster these results, means and purposes are assigned to scientific categories in the sci￾ence citation index and examples are presented below. Several ideas are identified that uses methods from ‘Artificial Intelligence’ (mean) for applications in ‘Health Care Sciences and Services’ (purpose). We also identify new ideas using ‘Imaging Science and Photographic Technology’ (mean) for ‘Medical Informatics’ purposes. Further ideas use techniques from ‘Remote Sensing’ (mean) in the field of ‘Tropical Medicine’ (purpose). Additionally, several ideas use ‘Computer Science, Theory and Methods’ (mean) for applications in ‘Psychiatry’ (purpose). Furthermore, methods from ‘Artificial Intelligence’ (mean) are used for ‘Automation and Control Systems’ purposes. To evaluate these results, we use precision and recall measures commonly used in information retrieval based on true positives, false positives and false negatives. For this, we have to define the ground truth for our evaluation. Therefore, a human expert also identifies new and useful ideas from these patents manually that means without using our idea mining approach. He uses the idea definition in Section 1.2. This means, he checks each text pattern for finding terms representing a known mean (purpose) and terms representing an unknown purpose (mean). These results are the ground truth for the evaluation. For each patent, we compute its precision and recall values by using the idea mining measure and by using the Jaccard’s coeffi- cient. Then, we compute the average precision and recall values. As a result, we get a precision value of 40% and a recall value of 25% by using the idea mining approach with the idea mining mea￾sure. A precision value of 40% means that if the idea mining ap￾proach extracts ten text patterns then four of them represent a new and useful idea. A recall value of 25% means that if there are four new and useful ideas in the new text then the idea mining ap￾proach extracts only one of them. In contrast to this, we get a pre￾cision value of 30% and a recall value of 20% by using Jaccard’s coefficient. This is because in some texts Jaccard’s coefficient ex￾tracts text patterns from the new text that are similar to text pat￾terns from the problem description. This represents probably a known idea but not a new idea. Beside Jaccard’s coefficient, we also test other well-known heu￾ristic measures like overlap-index, cosine-similarity and dice-sim￾ilarity (Ferber, 2003) as baseline. However, we get nearly the same results for the precision (30%) and for the recall (20%) value. 8. The idea mining application The idea mining application focus on users without extensive knowledge in the text mining field as well as on text mining ex￾perts. We give them the possibility to extract specifically problem solution ideas for their own needs using this idea mining approach. They can access to the web-based application via the internet. It is available under http://www.text-mining.info and it is programmed in perl and ruby. An user has to provide two textual files, a problem description and a new text that probably consists of problem solution ideas. These files can be formatted in various ways e.g. as plain text, html, xml etc. However, scripting code, (html- or xml-) tags, and images are discarded that means the application extracts plain text from the provided files. Then, the user has to select the language of these texts to integrate a general stop word list of this language. The application offers general stop word lists in English, German, Dutch, Spain and French. After determining the parameters of the application the automatically extraction of new and useful ideas from the new text starts as described in the idea mining process (see Sections 3 and 4). As a result, new ideas are presented as de￾scribed in Section 5. 9. Conclusions and future research This study shows the success of an automatic approach for find￾ing new ideas from textual information. For this, the study trans￾forms creativity approaches from psychology and cognitive science to text mining approaches. One main finding here is to redefine an abstract term (an idea) in a concrete way that it can be used for computing with text mining methods. In detail, it is shown that a technological idea represents a combination of a pur￾pose and a mean and that purposes and means are defined by a combination of terms, which co-occur. Additionally, it is shown that problems and problem solution ideas can be represented as term vectors in vector space model. For this, the study contributes a new (idea mining) measure. This measure identifies new ideas by comparing vectors that represent a problem to vectors that represent a problem solution idea. Last, it is shown that approaches from comprehensibility research can be adopted to this approach to present the new ideas in a comprehen￾sible way to the user. As further main finding, it is demonstrated that this theoretical approach can be realized by a web-based application. The success of the idea mining measure is proved by comparing it to further heuristic measures (overlap-index, co￾sine-similarity and dice-similarity). Directions for future research are given by the fact that nowa￾days there is a large amount of textual information available on the internet and this information probably contains many new technological ideas. Enlarging this approach to a web idea mining approach that automatically identifies problem solution ideas from the internet is an interesting topic for further research. Additionally, the parameters of the approach can be optimized and the idea mining measure can probably be enlarged with fur￾ther aspects to improve its quality that means to get better results for the precision and recall values. A further aspect is to transform this idea mining approach to the colloquial language. For this, it is necessary that the idea definition also contains new product ideas from the consumers. Then, new product ideas can be identified to support marketing activities. Last, the approach can be extended with innovation-related as￾pects. Then, extracted ideas can be classified as innovative ideas and might be used as starting point for the new product development. Acknowledgements We thank Joachim Schulze and Jörg Fenner for constructive technical comments. References Coussement, K., & Van den Poel, D. (2008). Integrating the voice of customers through call center emails into a decision support system for churn prediction. Information and Management, 45, 165. Coussement, K., & Van den Poel, D. (2009). Improving customer attrition prediction by integrating emotions from client/company interaction emails and evaluating multiple classifiers. Expert Systems with Applications, 36, 6127–6134. D. Thorleuchter et al. / Expert Systems with Applications 37 (2010) 7182–7188 7187

D. Thorleuchter et al. Expert Systems with Applications 37(2010)7182-7188 Dagdeviren, M, Yavuz, S,& Kilinc, N (2009). Weapon selection using the AHP and Porter, M. F(1980).An m for suffix stripping. Program, 130-137. PSiS methods under fuzzy environment. Expert Systems with Applications, 36, Puppe, F Stoyan, H, Studer, R(2003) Fenner,J& Thorleuchter, D (2009) Textmining-Analyse von Forschungsvorhaben P61】MmDg4 ndbuch der Kainstlichen Intelli des National Institute of Standards and Technology. Euskirchen: Fraunhofer INT teller objekte der Forschungsforderung. In H. Paschen H. Krauch Ferber, R (2003). Information retrieval. Heidelberg: dpunkt verlag. pp 74-80 esch, R(1948) A new readability yardstick. Journal of Applied Psychology, 32, (Eds ) Methoden und probleme der Forschungs- und Entwicklungsplanung(pp. 47). Rohpohl, G.(1996). Das Ende der Natur. In L Schafer E. Straker (Eds ) Gentsch, P,& Hanlein, M. (1999). Text mining. WSU. 12, 1646. er, D.(2008). Finding technological ideas and inventions with text for special purpo gy research. Berlin: Walt Tho ., Prinze, A (2010a). A compared R& D-based and Hotho, A, Numberger, A& Paak, G(2005). A brief survey of text mining. LDV patent-based cross impact analysis for identifying relationships between technologies. Technological Forecasting and Social Change, doi: 10.1016/ Langer, L, Schulz V. Thun, F,& Tausch, R.(1974). Verstandlichkeit in Schule und Thorleuchter, D, Van den Poel, D, Prinze, A(2010b) Extracting consumers needs Li, Y R, Wang. L H, Hong C F (2009). Extracting the significant-rare keywords for patent analysis. Expert Systems with Applications, 36, 5200-5204. gMD040知 lamine n m Gon&QLu(E Martin-Bautista, M. J- Sanches. D, Serrano, M,& vila, M. A (2004) Text mining Thorleuchter, D, Van den Poel, D Prinzie, A(201 ing fuzzy association rules In V Loia, M. Nikravesh, &LA Zadeh(Eds Fuzzy ent. In H. Locarek-Junge, C. logic and the internet(pp. 173). Berlin: Springer-Verlag. Weihs(Eds ). Classification as a tool for research. Berlin: Springer-verla Osborn, A-F. (1948) Your creative power. New York: C. Scribner,s sons. Wallas, G (1926). The art of thought. New York: Harcourt Brace

Dagdeviren, M., Yavuz, S., & Kilinc, N. (2009). Weapon selection using the AHP and TOPSIS methods under fuzzy environment. Expert Systems with Applications, 36, 8150. Fenner, J., & Thorleuchter, D. (2009). Textmining-Analyse von Forschungsvorhaben des National Institute of Standards and Technology. Euskirchen: Fraunhofer INT Edition. Ferber, R. (2003). Information retrieval. Heidelberg: dpunkt.verlag. pp. 74–80. Flesch, R. (1948). A new readability yardstick. Journal of Applied Psychology, 32, 221–233. Gentsch, P., & Hänlein, M. (1999). Text mining. WISU, 12, 1646. Groeben, N. (1982). Leserpsychologie: Textverständnis – Textverständlichkeit. Münster: Aschendorff. Hoffmann, L., Kalverkämper, H., & Wiegand, H. E. (1998). Fachsprachen – Languages for special purposes: Ein internationales Handbuch zur Fachsprachenforschung und Terminologiewissenschaft – An international handbook of special-language and terminology research. Berlin: Walter de Gruyter. p. 1602. Hotho, A., Nürnberger, A., & Paaß, G. (2005). A brief survey of text mining. LDV Forum, 20(1), 19–26. Langer, I., Schulz v. Thun, F., & Tausch, R. (1974). Verständlichkeit in Schule und Verwaltung. München: Ernst Reinhardt. Li, Y. R., Wang, L. H., & Hong, C. F. (2009). Extracting the significant-rare keywords for patent analysis. Expert Systems with Applications, 36, 5200–5204. Martin-Bautista, M. J., Sanches, D., Serrano, J. M., & Vila, M. A. (2004). Text mining using fuzzy association rules. In V. Loia, M. Nikravesh, & L. A. Zadeh (Eds.), Fuzzy logic and the internet (pp. 173). Berlin: Springer-Verlag. Osborn, A.-F. (1948). Your creative power. New York: C. Scribner’s sons. Porter, M. F. (1980). An algorithm for suffix stripping. Program, 14(3), 130–137. Puppe, F., Stoyan, H., & Studer, R. (2003). Knowledge engineering. In G. Görz, C. R. Rollinger, & J. Schneeberger (Eds.), Handbuch der Künstlichen Intelligenz (pp. 611). München: Oldenbourg. Ripke, M., & Stöber, G. (1972). Probleme und Methoden der Identifizierung potentieller Objekte der Forschungsförderung. In H. Paschen & H. Krauch (Eds.), Methoden und Probleme der Forschungs- und Entwicklungsplanung (pp. 47). München: Oldenbourg. Rohpohl, G. (1996). Das Ende der Natur. In L. Schäfer & E. Sträker (Eds.), Naturauffassungen in Philosophie, Wissenschaft und Technik, Freiburg (pp. 143–163). München: Alber. Thorleuchter, D. (2008). Finding technological ideas and inventions with text mining and technique philosophy. In C. Preisach, H. Burkhardt, L. Schmidt￾Thieme, & R. Decker (Eds.), Data analysis, machine learning, and applications (pp. 413–420). Berlin: Springer-Verlag. Thorleuchter, D., Van den Poel, D., & Prinzie, A. (2010a). A compared R&D-based and patent-based cross impact analysis for identifying relationships between technologies. Technological Forecasting and Social Change, doi:10.1016/ j.techfore.2010.03.002. Thorleuchter, D., Van den Poel, D., & Prinzie, A. (2010b). Extracting consumers needs for new products - a web mining approach. In M. Gong, & Q. Luo (Eds.), Proceedings WKDD 2010 (pp. 440). Los Alamitos: IEEE Computer Society. Thorleuchter, D., Van den Poel, D., & Prinzie, A. (2010c). Mining innovative ideas to support new product research and development. In H. Locarek-Junge, & C. Weihs (Eds.), Classification as a tool for research. Berlin: Springer-Verlag. Wallas, G. (1926). The art of thought. New York: Harcourt Brace. 7188 D. Thorleuchter et al. / Expert Systems with Applications 37 (2010) 7182–7188