Welcome to the repository of the Data Science Capstone project. The shiny app is fully sourced here, and the main preparation work is under "data_science_capstone.R"
http://calin.shinyapps.io/predict_next_word/
http://rpubs.com/calin/predict_next_word
Overall top-3 score: 2.97 % Overall top-1 precision: 2.71 % Overall top-3 precision: 3.12 % Average runtime: 3597.09 msec Number of predictions: 1808 Total memory used: 1228.82 MB
Dataset details Dataset "blogs" (50 lines, 999 words, hash be3b52b1eab7c5c82708d9000c646a7a179ff9eac01164eff0de2b8bbea3d3c9) Score: 4.39 %, Top-1 precision: 3.95 %, Top-3 precision: 4.66 % Dataset "tweets" (50 lines, 830 words, hash 020a348a8f49d0754ea10e8e4ff78a98d5e99bab7482fd9340bcfcea4a5c4dbc) Score: 1.54 %, Top-1 precision: 1.46 %, Top-3 precision: 1.59 %