Data Mining Algorithms In R/Packages/arules/adult

Case Study

Analysis

We have observed that CSPADE found many trivial sequences from user behaviour. For example, it has found many unitary sequences, such as <{design}>, <{ajax}>, <{css}>, among others. These unitary sequences are really frequently used, but they may not be useful in the particular application, which is Tag Recommendation.

Furthermore, other trivial sequences were found, such as <{design}.{design}> and <{webdesign},{design}>. These sequences indicates that the same users tend to bookmarks pages in the same subject subsequently. However, some interesting patters were also found. We can cite <{library},{books}>, <{javascript},{ajax}> and <{video},{youtube}>.

We can also observe that many frequent patterns are related to design, art and web_development. These tags are also the most popular tags in the whole Delicious system, as can be seen here.

References

1. ^ Li, K.-C., Chang, D.-J., Rouchka, E. C., Chen, Y. Y., 2007. "Biological sequence mining using plausible neural network and its application to exon/intron boundaries prediction". In: CIBCB. IEEE, pp. 165–169.
2. ^ Peng, W.-C., Liao, Z.-X., 2009. "Mining sequential patterns across multiple sequence databases". Data Knowl. Eng. 68 (10), 1014–1033.
3. ^ Telecom Paper, January 2009. "Google query volume".
4. ^ Zaki, M. J., 2001. "Spade: An efficient algorithm for mining frequent sequences". In: Machine Learning. Vol. 42. pp. 31–60.