Mining unexpected patterns using decision trees and interestingness measures: a case study of endometriosis

Because clinical research is carried out in complex environments, prior domain knowledge, constraints, and expert knowledge can enhance the capabilities and performance of data mining. In this paper we propose an unexpected pattern mining model that uses decision trees to compare recovery rates of two different treatments, and to find patterns that contrast with the prior knowledge of domain users. In the proposed model we define interestingness measures to determine whether the patterns found are interesting to the domain. By applying the concept of domain-driven data mining, we repeatedly utilize decision trees and interestingness measures in a closed-loop, in-depth mining process to find unexpected and interesting patterns. We use retrospective data from transvaginal ultrasound-guided aspirations to show that the proposed model can successfully compare different treatments using a decision tree, which is a new usage of that tool. We believe that unexpected, interesting patterns may provide clinical researchers with different perspectives for future research. Similar content being viewed by others Abbreviations - ANOVA: - Analysis of variance - D\(^{3}\)M: - Domain-driven data mining - CA-125: - Cancer antigen 125, carcinoma antigen 125, or carbohydrate antigen 125 - BMI: - Body mass index - CART: - Classification and regression tree - EST: - Ethanol sclerotherapy - ID3: - Iterative Dichotomiser 3 algorithm - CHAID: - Chi-Square Automatic Interaction Detector - tech(): - Technical interestingness measures - biz(): - Business interestingness measures - act(): - Actionability of a pattern - tech_obj(): - Technical objective interestingness measures - tech_sub(): - Technical subjective interestingness measures - biz_obj(): - Business objective interestingness measures - biz_sub(): - Business subjective interestingness measures - RecoveryRate(): - Probability patient recovers from illness - IM \(_{\mathrm{tech\_obj}}\) : - Technical objective interestingness measure - IM \(_{\mathrm{tech\_sub}}\) : - Technical subjective interestingness measure - IM \(_{\mathrm{biz\_obj}}\) : - Business objective interestingness measure - IM \(_{\mathrm{biz\_sub}}\) : - Business subjective interestingness measure

Baena-García M, Morales-Bueno R (2012) Mining interestingness measures for string pattern mining. Knowl-Based Syst 25:45–50. doi:10.1016/j.knosys.2011.01.013 Bay SD, Pazzani MJ (2001) Detecting group differences: mining contrast sets. Data Min Knowl Disc 5:213–246. doi:10.1023/a:1011429418057 Berlanda N, Vercellini P, Fedele L (2010) The outcomes of repeat surgery for recurrent symptomatic endometriosis. Curr Opin Obstet Gynecol 22:320–325 Bolton S, Bon C (2009a) Analysis of variance. Pharmaceutical statistics: practical and clinical applications, 5th edn. Informa Healthcare, New York, pp 182–221 Bolton S, Bon C (2009b) Linear regression and correlation. Pharmaceutical statistics: practical and clinical applications, 5th edn. Informa Healthcare, New York, pp 147–181 Breiman L (1984) Classification and regression trees. In: The Wadsworth statistics/probability series. Wadsworth International Group, Belmont Bulletti C, Coccia M, Battistoni S, Borini A (2010) Endometriosis and infertility. J Assist Reprod Genet 27:441–447 Cao L, Zhang C (2007) Domain-driven, actionable knowledge discovery. IEEE Intell Syst 22:78–88 Cao L, Luo D, Zhang C (2007) Knowledge actionability: satisfying technical and business interestingness. Int J Bus Intell Data Min 2:496–514. doi:10.1504/ijbidm.2007.016385 Cao L, Zhang C, Yu PS, Zhao Y (2010a) Challenges and trends. Domain driven data mining. Springer, US, pp 1–25 Cao L, Zhang C, Yu PS, Zhao Y (2010b) D\(^3\)M methodology. Domain driven data mining. Springer, US, pp 27–47 Cao L, Zhang C (2006) Domain-driven actionable knowledge discovery in the real world. In: Ng W-K, Kitsuregawa M, Li J, Chang K (eds) Advances in knowledge discovery and data mining, Lecture notes in computer science, vol 3918. Springer, Berlin, pp 821–830. doi:10.1007/11731139_96 Chawla NV, Japkowicz N, Kotcz A (2004) Editorial: special issue on learning from imbalanced data sets. SIGKDD Explor Newsl 6:1–6. doi:10.1145/1007730.1007733 Donnez J, Squiffle J, Donnez O (2011) Minimally invasive gynecologic procedures. Curr Opin Obstet Gynecol 23:289–295. doi:10.1097/GCO.0b013e328348a283 Freitas AA (1999) On rule interestingness measures. Knowl-Based Syst 12:309–315 Geng L, Hamilton HJ (2006) Interestingness measures for data mining: a survey. ACM Comput Surv 38:1–31. doi:10.1145/1132960.1132963 Glass DH (2013) Confirmation measures of association rule interestingness. Knowl-Based Syst 44:65–77. doi:10.1016/j.knosys.2013.01.021 Hsieh C-L, Shiau C-S, Lo L-M, Hsieh Ts-Ta, Chang M-Y (2009) Effectiveness of ultrasound-guided aspiration and sclerotherapy with 95 % ethanol for treatment of recurrent ovarian endometriomas. Fertil Steril 91:2709–2713 Ikuta A et al (2006) Management of transvaginal ultrasound-guided absolute ethanol sclerotherapy for ovarian endometriotic cysts. J Med Ultrason 33:99–103 Kafali H, Yurtseven S, Atmaca F, Ozardali I (2003) Management of non-neoplastic ovarian cysts with sclerotherapy. Int J Gynaecol Obstet 81:41–45 Kass GV (1980) An exploratory technique for investigating large quantities of categorical data. Appl Stat 29(2):119–127 Kennedy S et al (2005) ESHRE guideline for the diagnosis and treatment of endometriosis. Hum Reprod 20:2698–2704 Kontonasios K-N, Spyropoulou E, De Bie T (2012) Knowledge discovery interestingness measures based on unexpectedness. Wiley Interdiscip Rev Data Min Knowl Discov 2:386–399 Lenca P, Meyer P, Vaillant B, Lallich S (2008) On selecting interestingness measures for association rules: user oriented description and multiple criteria decision aid. Eur J Oper Res 184:610–626. doi:10.1016/j.ejor.2006.10.059 Ling CX, Tielin C, Qiang Y, Jie C (2002) Mining optimal actions for profitable CRM. In: Paper presented at the proceedings of the 2002 IEEE international conference on data mining, 2002 Liu B, Hsu W, Mun L-F, Lee H-Y (1999) Finding interesting patterns using user expectations. IEEE Trans Knowl Data Eng 11:817–832. doi:10.1109/69.824588 McGarry K (2005) A survey of interestingness measures for knowledge discovery. Knowl Eng Rev 20:39–61. doi:10.1017/s0269888905000408 Nap AW, Groothuis PG, Demir AY, Evers JLH, Dunselman GAJ (2004) Pathogenesis of endometriosis. Best Pract Res Clin Obstet Gynaecol 18:233–244 Noma J, Yoshida N (2001) Efficacy of ethanol sclerotherapy for ovarian endometriomas. Int J Gynaecol Obstet 72:35–39 Padmanabhan B, Tuzhilin A (1999) Unexpectedness as a measure of interestingness in knowledge discovery. Decis Support Syst 27:303–318 Piatetsky-Shapiro G (1991) Discovery, analysis, and presentation of strong rules. In: Piatetsky-Shapiro G, Frawley W (eds) Knowledge discovery in databases. AAAI/MIT Press, Cambridge, pp 229–248 Quinlan JR (1993) C4.5: programs for machine learning. Morgan Kaufmann Publishers Inc, San Francisco Quinlan JR (1986) Induction of decision trees. Mach Learn 1:81–106. doi:10.1007/bf00116251 Rokach L, Maimon O (2008) Data mining with decision trees: theory and applications. World Scientific Publishing Company, MA Sebastian Y, Then PHH (2011) Domain-driven KDD for mining functionally novel rules and linking disjoint medical hypotheses. Knowl-Based Syst 24:609–620 Shaharanee INM, Hadzic F, Dillon TS (2011) Interestingness measures for association rules based on statistical validity. Knowl-Based Syst 24:386–392. doi:10.1016/j.knosys.2010.11.005 Silberschatz A, Tuzhilin A (1995) On subjective measures of interestingness in knowledge discovery. In: Paper presented at the proceedings of the 1st international conference on knowledge discovery and data mining (KDD’ 95) Tsay L-S, Raś ZW (2005) Action rules discovery: system DEAR2, method and experiments. J Exp Theory Artif Intell 17:119–128 Vercellini P, Somigliana E, ViganÒ P, De Matteis S, Barbara G, Fedele L (2009) The effect of second-line surgery on reproductive performance of women with recurrent endometriosis: a systematic review. Acta Obstet Gynecol Scand 88:1074–1082. doi:10.1080/00016340903214973 Wang YF, Chang MY, Chiang RD, Hwang LJ, Lee CM, Wang YH (2013) Mining medical data: a case study of endometriosis. J Med Syst 37:1–7. doi:10.1007/s10916-012-9899-y Wang K, Zhou S, Han J (2002) Profit mining: from patterns to actions. In: Paper presented at the proceedings of the 8th international conference on extending database technology: advances in database technology Yao Y, Chen Y, Yang X (2006) A measurement-theoretic foundation of rule interestingness evaluation. In: Young Lin T, Ohsuga S, Liau C-J, Hu X (eds) Foundations and novel approaches in data mining, Studies in computational intelligence, vol 9. Springer, Berlin, pp 41–59. doi:10.1007/11539827_3 Zhu Z, Gu J, Zhang L, Song W, Gao R (2009) Research on domain-driven actionable knowledge discovery. In: Shi Y, Wang S, Peng Y, Li J, Zeng Y (eds) Cutting-edge research topics on multiple criteria decision making, Communications in computer and information science, vol 35. Springer, Berlin, pp 176–183. doi:10.1007/978-3-642-02298-2_27 Zhu W, Tan Z, Fu Z, Li X, Chen X, Zhou Y (2011) Repeat transvaginal ultrasound-guided aspiration of ovarian endometrioma in infertile women with endometriosis. Am J Obstet Gynecol 204:61.e61–61.e66 Author information Authors and Affiliations Corresponding author Additional information Communicated by V. Loia. Rights and permissions About this article Cite this article Chang, MY., Chiang, RD., Wu, SJ. et al. Mining unexpected patterns using decision trees and interestingness measures: a case study of endometriosis. Soft Comput 20, 3991–4003 (2016). https://doi.org/10.1007/s00500-015-1735-0 Published: Issue date: DOI: https://doi.org/10.1007/s00500-015-1735-0