Web Annotation becomes an important collaboration technique for users of the web. In order to increase universal collaboration, it is vital for annotators to contact the most related people that share the same interests. Although annotations themselves can be used to exchange ideas and experts, the improper writing of their attached notes could decrease the intended collaboration due to the lack of expressing ideas clearly. Adorning annotations with proper tags makes it easier for annotators to express their feelings, emotions, and interests through their submitted annotations. The ability to use the tags in searching for annotations leads to reaching the most related tags to one's interests and this indeed increases changing ideas between annotators and hence their collaboration. However, expressing the ideas in annotations with proper tags is not an easy task for most annotators. On the opposite side, attaching improper tags loses the soul and the intention of creating annotations which decreases the amount of collaboration. This work is related to automating the process of tags' suggestion by studying the texts selected by annotators over the web. Suggesting proper tags takes into account the part of speech of the annotated texts. The experimental results conducted in the work are related to finding the most suitable threshold for the percentage number of tag suggestions.