CS 412 Intro to Data Mining Chapter 13

CS 412 Intro. to Data Mining Chapter 13. Trends and Research Frontier in DM Qi Li, Computer Science, Univ. Illinois at Urbana-Champaign, 2019 1

Chapter 13. Trends and Research Frontier in DM 2 q Complex data types q Applications q Data mining competitions q Ethical issues of data mining

Complex Data Types 3 q Sequences q Graphs q Text q Web q Stream, spatiotemporal, multimedia, Io. T…

Sequences Time Series Data (e. g. , stock market data) q Symbolic sequences (e. g. , customer shopping sequences, web clickstreams) q Biological sequences (e. g. , DNA sequences) q 4

Graph Homogeneous graph (nodes/links are of same type) q Heterogeneous (nodes/links are of different types) q 5

Text q 6 Text mining is an interdisciplinary field that draws on information retrieval, data mining, machine learning, statistics, and computational linguistics. A substantial portion of information is stored as text, such as news articles, technical papers, books, digital libraries, e-mail messages, blogs, and Web pages. Hence, research in text mining has been very active, an important goal of which is to derive high-quality information from text.

Web content q Web structure q Web usage q 7

Challenges q 8

Chapter 13. Trends and Research Frontier in DM 9 q Complex data types q Applications q Data mining competitions q Ethical issues of data mining

Recommender System q q q Product recommendation (Amazon, EBay) Search recommendation (Google, Bing) Video/music/post recommendation (Netflix, Pandora, Pinterst) Friend recommendation (Facebook, twitter) Job recommendation (link. In) collaborative filtering q content-based filtering q hybrid q 10

Commerce, Profiling and Finance Planning and Forecasting q Dynamic pricing q Ads bidding q Profiling q User profiling q Churn Prediction: knowing which users are going to stop using your platform in the future. q Product profiling q Fintech q Stock market q Sentiment analysis q 11

Urban Planning Energy and power q Traffic prediction and management q Parking detection q Traffic control q Transportation sharing system q Uber q Bike-sharing q Pollution q Air quality prediction q 12

Medicine and Healthcare Disease prediction q Computer Aided Detection q EHR q Risk prediction q Disease progression prediction q Healthcare q Epidemic and outbreak prediction q Food safety q Medicine study q Drug discovery and prediction q Bioinformatics q 13

Other Sciences and Applications Education q MOOC (massive open online course) q Political science and Social science q Fake news q Crime and terrorist detection q Disaster detection q Opinion mining q Social influence q Environmental Science q Climate q 14

Chapter 13. Trends and Research Frontier in DM 15 q Complex data types q Applications q Data mining competitions q Ethical issues of data mining

Data mining competitions KDD cup q https: //www. kdd. org/kdd 2019/kdd-cup q WSDM cup q http: //www. wsdm-conference. org/2018/call-for-participants. html q ICDM Contest q http: //icdm 2019. bigke. org/ q 16

KDD Cup q q q q 17 KDD cup 2018: forecast air quality KDD Cup 2017: Highway Tollgates Traffic Flow Prediction KDD Cup 2016: Whose papers are accepted the most: towards measuring the impact of research institutions KDD Cup 2015: predicting students’ likelihood of dropout on MOOC KDD Cup 2014: Predict funding requests that deserve an A+ KDD Cup 2013 (Track 2): Identify which authors correspond to the same person KDD Cup 2013 (Track 1): Determine whether an author has written a given paper

Chapter 13. Trends and Research Frontier in DM 18 q Complex data types q Applications q Data mining competitions q Ethical issues of data mining

Ethical Issues of Data Mining Privacy and safety q Information reliability q Information Bias q Expandability q 19