340 30 1 SP2-N 2015 Onomatoperori : Ranking Cooking Recipes by using Onomatopoeias which Express their Tastes and Textures Chiemi Watanabe Satoshi Nakamura Graduate School of Systems and Information Engineering, University of Tsukuba chiemi@cs.tsukuba.ac.jp, http://www.watalab.net/ chiemi/, JST CREST School of Interdisciplinary Mathematical Sciences, Meiji University, JST CREST satoshi@snakamura.org, http://snakamura.org/ keywords: Recipe retrieval, Ranking algorithm, Co-occurrence factor, sentiment search Summary This paper proposes a ranking methodology of cooking recipe by using fitness value between a recipe and onomatopoeia. This system is implemented as a function of a cooking recipe search site Onomatoperori. By using onomatopoeia, users can find what they want to cook from their ambiguous idea. We defined formulas for calculating fitness value between recipe and onomatopoeia by using mutual information between onomatopoeia and a word in title or description of recipes. In addition, we defined the similarity measure between onomatopeia words by mapping their words by using 15 sentimental dimensions for expressing the tastes and textures of the dishes. And we improve the ranking methodology by using the similarity among onomatopoeia words. By using these ranking methodologies we can search the cooking recipes which are related to the onomatopoeia although they do not include the onomatopeia word in the recipes. 1. Cookpad 73259 2 1.
341 2. 15 [ 12][Komatsu 12][ 14][ 14][ 11] 3. 1 [Marchionini 06][ 10] 2. SQL PHP Ajax 1 ( 1(1)) 1(2) ( 1(3)) 1(2) ( 1(3)) 1(3) -- 1 ( 1(4)) 2 3 5 3. 2 3 4 5 6 (1) (2) 1
342 30 1 SP2-N 2015 1 2 10 0.00350 0.00247 0.00242 0.00226 0.00180 0.00155 0.00122 0.00099 0.00094 0.00094 1,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, (3) (2) 3 1 (1) (2) [ 03] 126 ( 1 252 3 2 3 10 0.00290 0.00272 0.00230 0.00208 0.00180 0.00161 0.0015 0.00155 0.00155 0.00155 1 2 o w Col(o,w) Col(o,w)=p(o,w)log p(o,w) p(o)p(w) 10 2 3
343 4 3 2!!!!!!!! 2 2!!!?!!! 3 3 o r Fitness(o,r) Fitness(o,r)=Σ w Wr Col(o,w) 4 3 10 5 3 5 1 5 10 4 3 10 5 10 1 0.00175 2 0.00138 3 0.00131 4 0.00112 5 0.00104 6!!! 0.00098 7 2 0.00097 8 2 0.00089 9?? 0.00088 10 Coffee 0.00085 6 20 4. 3
344 30 1 SP2-N 2015 15 4 1 [ 03] 37 15 (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) 2 crowd4u 4 2 252 15 1 2 10 252 15 10 10.5 1 Crowd4U[Morishima 12] Crowd4U 5 Crowd4U Crowd4U 5 Crowd4U 2 3 Crowd4U 252 15 10 37800 15 10
345 3 4 10 10 10 o v o 15 v o =(o figure,o density,o transparent, o elastic,o temperature,o hard, o weight,o smooth,o dry,o sticky, o comfort,o hot,o sweet,o salty,o sour ) A 2 2 15 GOROGORO 1 4 0 5. r v r =(r 1,...,r 15 ) r o Onomatope v o =(o 1,...,o 15 ) Fitness(o,r) v r =Σ o Onomatope Fitness(o,r)v o o r Fitness sentiment (o,r)= v r v o v r v o 5 r
346 30 1 SP2-N 2015 7 10 0.90794411 0.89740909?? 0.89168952 0.88264412 0.86773076 0.86751795 0.86364215 0.86355409 0.86335166? 0.86260126 5 Fitness(,r)=0.85 Fitness(,r)=0.63 Fitness(,r)=0.01 5 r r 10 7 5 8 6. 6 1 3 ID 3 15 8 4 4 8 4 8 1 1 10 19 21 4 6
347 9 ONO 0.663 0.828 0.824 1.450 0.721 0.887 0.793 1.383 0.596 0.769 0.853 1.5235 6 3 http://perori.nkmr.io/ex 1 10 3 30 30 3 3 Cookpad 2 73259 17708 24% 8 6 10 2 http://www.cookpad.com 2 8 6 2 9 2, 1 0 ( 9 ONO ONO 6 2 1 ONO 77% 8 7 6 1 2 8 9
348 30 1 SP2-N 2015 7 10 10 5 0.00596629 0.00415195 0.00385264 0.00354592 0.00092016 8 10 (1) (2) 1 A.1 9 2 10 5 11 5 4 2 (1) (2)
349 11 11 5 0.08323364 0.04202669 0.00663583 0.00596629 0.00541931 o 1 r o 1 r o 2 Fitness(r, o 2 ) o 1 Fitness(r, o 1 ) o 1 r 11 0.15 0.6 3 0.1 7. [ 14] Twitter [Kato 12] LDA 1 1 [ 14] [ 10] [ 14] [ 12] [Komatsu 12] [ 14] [ 11] 2 [ 14], [ 14] [ 14]
350 30 1 SP2-N 2015 Exploratory Search [Marchionini 06][ 10] [ 10] [ 11] [ 10] [ 12] [ 12] SIFT [ 13] 8. 15 [ 03] [ 05] [ 03] Crowd4U Crowd4U (A) 23680006)
351 [Kato 12] Kato, A.: Discovering pattern of onomatopoeia used in food reviews, in Soft Computing and Intelligent Systems (SCIS) and 13th International Symposium on Advanced Intelligent Systems (ISIS), pp. 502 507 (2012) [Komatsu 12] Komatsu, T.: Quantifying Japanese Onomatopoeias: Toward Augmenting Creative Activities with Onomatopoeias, in the 3rd Augmented Human International Conference (AH12), No. a-15 (2012) [Marchionini 06] Marchionini, G.: Exploratory search: from finding to understanding, in Communications of ACM, Vol. 49, pp. 41 46 (2006) [Morishima 12] Morishima, A., Shinagawa, N., Mitsuishi, T., Aoki, H., and Fukusumi, S.: CyLog/Crowd4U: A Declarative Platform for Complex Data-centric Crowdsourcing, in PVLDB, pp. 1918 1921 (2012) [ 14],,, Twitter, 6 (DEIM2014), pp. B6 6 (2014) [ 11],,,,, Vol. 11, No. 4, pp. 545 552 (2011) [ 10],,, Web, (DBSJ Journal), Vol. 9, No. 1, pp. 41 46 (2010) [ 14],,,,, 6 (DEIM2014), pp. P2 3 (2014) [ 10],, RerankEverything,, Vol.3, No. 4, pp. 48 64 (2010) [ 11],, Q&A,, Vol. 4, No. 2, pp. 74 87 (2011) [ 12], :, 26, pp. 1M2 OS 8b 2 (2012) [ 12],,, Vol. J92-A, No. 11, pp. 752 763 (2012) [ 10],,, Vol. 27, No. 1 (2010) [ 03], (2003) [ 14],,,, Vol. 29, No. 1, pp. 41 52 (2014) [ 05],,,,,,,, Vol. 52, No. 8, pp. 337 346 (2005) [ 14],,,, Vol. 15, No. 3, pp. 487 490 (2014) [ 14],,,, Vol. 13, No. 2, pp. 353 359 (2014) [ 13],,,,, 27, No. 3N1-OS-01a-1, pp. 1 4 (2013) 1975 2003 2005 2013 A. 1976 2004 2006 2009 2013 2014 04 28
352 30 1 SP2-N 2015 A.1