As table summarizations naturally plays a great role in big data analysis, the profit of using these algorithms has to be considered. Knowledge gain and limitations of using the algorithms needs to be evaluated. Given the limited capability of people to process data, which is believed to be 7+-2 items for short term memory, and the ever growing amount of data, it is vital to provide techniques, that can efficiently reduce the cardinality of the initial dataset. Summaries then can be processed by humans more easily, which provides easy access to the dataset in question and underlying patterns of interest. Patterns that provide useful information about the dataset must be the ones, that cover a big portion of the dataset thus making them contain frequent values in the dataset. There are some areas discussed in the papers which benefit from these techniques. For example the possibility to rank and pre-process web tables in order to make them more applicable for a user search query. Another application for summarization lies in the field of security. As mentioned in one paper, it is impossible for an operator to sight all internet traffic manually even in small networks. To detect possible intruders operators have to take action quickly and eliminate the threat to the network. Summarization provides the possibility to detect malicious traffic, by visualizing suspicious traffic in a summary. A possibility in the field of security yet to explore, could be the application of summarization techniques in virus, malware and spyware scans. It might be possible to achieve better results in some cases by employing summarization techniques there. Reducing the cardinality by applying summarization algorithms hence reduces the required searchtime. Also it provides the possibility to determine the relevance of a dataset with a glance, for both man and machine alike. Making it possible for the user to see the outline of a dataset immediately. Another benefit of summaries is, that it shows relations between different attribute values of tuples. The shown benefits hold true for homogeneous datasets however and there might be datasets where values differ too much to be summarized effectively. In such cases summarization can be difficult.