Research Article

Two-Level Automatic Adaptation of a Distributed User Profile for Personalized News Content Delivery

Table 3

%Error for the exact ranking position of each article in comparison to the manual ranking per user (25 users).

User12345 25

10%63.16%15.79%15.79%21.05% 5.26%
Articles25.56%22.22%5.56%27,28%0% 11.11%
ranking30%41.18%23.53%82.35%0% 17.65%
position40%37.50%12.50%6.25%18.75% 12.50%
56.67%6.67%13.33%6.67%6.67% 13.33%
Short-term
Learning166.67%13.33%20%33.33%13.33% 26.67%
using1737.50%18.75%6.25%18.75%12.50% 12.50%
nouns & NEs1811.76%17.65%11.76%11.76%11.76% 11.76%
(Formula (3), (4))195.56%0%44.44%0%11.11% 16.67%
2021.05%21.05%15.79%31.58%47.37% 10.53%

Average16.69%24.79%20.01%28.35%23.10% 17.23%

110.53%44.44%62.50%26.67%33.33% 83.33%
Articles233.33%27.28%6.67%12.50%34.50% 67.40%
ranking341.18%55.70%39.20%37.50%33.33% 34.50%
position462.50%22.30%58.67%34.50%22.22% 17.65%
513.33%10.53%12.50%33.33%6.67% 42.80%
Short-term
Learning166.67%10.53%39.20%59.70%27.28% 37.50%
using only1793.75%44.44%27.28%55.70%6.67% 22.22%
nouns18100%12.50%28.35%44.44%26.67% 17.23%
(Formula (3))1983.33%23.10%31.58%13.33%58.67% 12.50%
2036.84%83.33%21.05%28.35%10.53% 39.20%

Average46.60%33.33%29.20%41.18%25.75% 44.44%