Understanding the story behind the most intriguing historical events and the influence of particular people and institutions on their course of action is a very difficult and time consuming process which requires a thorough study of a wide variety of historical records. Having this in mind, the main idea of our project is to find a constructive method to interactively visualize historical data, namely the various connections between the most prominent individuals in Venice history and their families in a clear, interesting and intuitive way. In this paper, we explain how we gathered, processed and structured the relevant data for our project. Additionally we discuss different possibilities for visually displaying our database and give an insight into the potential of such a tool.
First step in the realization of this project was finding the reliable Data sources. We needed an assortment of documents, texts and descriptions that would contain reliable information about important people of the Venetian social and political life. The first important find was a list of all the important Venetian families from 7th to 18th century that we found on Wikipedia. However, this list covers a very large time span and the available information varies dramatically both qualitatively and quantitatively between different families. Consequently, we have decided to focus our search on the doge related families as they were the most important and well documented. Resulting from this new query, we have found a list of all the doges of Venice in chronological order in Wikipedia. List contains the chronological Dodge Succession from 7th to 18th century. Again, some Doges were very well documented whereas the information about the others was scarce. Consequently, we decided to concentrate on the Doge related families following the Dodge succession over the course of 12th and 13th century as they were more vividly documented.
Understanding Data/ Data properties
Furthermore, having in mind that the main focus of our project was on displaying information related to social interactions of most prominent members of ancient Venetian community, the initial idea was to build a facebook-like database. However, considering that our research is mostly based on Wikipedia and some of the historical records that were referenced there, the emphasis is mostly put on important historical events and the relations between people in terms of their participation in these events along with their political and economic influences. Unfortunately, we did not manage to find any information about the, in true sense, social interaction between these characters such as letters or contracts. Even the information related to the family members is mostly oriented towards important connections with other prominent historical characters achieved through marriage or other interactions, e.g. Enrico Dandolo’s granddaughter, Ana Dandolo, was married to the Serbian King, Stefan Nemanjic.
Methods for data visualization
Taking into account that the quality of available data is in the rich information about the connections and not the social interactions, it is obvious that a facebook-like database is not appropriate for presenting the data of this type as it is made to facilitate the storyline and social interactions and not to provide a clear and intuitive visualization of the connections between people. On the other hand, we were using simple graphs as intermediary means to help us to better understand the connections between these historical figures and they turned out to be very useful as they inherently display the interconnections and give insight into the level of coupling of every individual with the rest of the population which is exactly what is needed. Consequently, we have visualized our database in form of a graph where each node corresponds to a historical character and each edge to a connection between two people.
Data Structuring/Technical means
However, we still had to choose an appropriate format to build our data base so that it can be processed by some automatic tool in order to visualize it in form of a graph. For this use, we have considered three standard formats XML, RDF and GraphML. Using the standard formats enables future development and easy integration with other projects built on similar standards. From the three mentioned standards, we have decided to build our database using GraphML as it is a widely used and more importantly standard format. Furthermore, in contrast to XML and RDF, GraphML is built especially for graphs and there are many ready-to-use tools for processing data in this format like Gephi, which is the free software that we utilized to construct the graphs.
Data Handling and Filtering
After finally concluded about the format and the visualization tool we were going to use, we dedicated ourselves to the completion of the database and familiarizing with the new software. They were both not easy tasks. After we have gathered all the doges and the persons related to them as well as useful background information about the nature of these connections, we had to decide what kind of attributes it would be appropriate to assign to the characters. Gephi supported attributes for both the characters and the connections. Considering the acquired information and the remaining time at that point, as well as the technical difficulty of the implementation, we concluded that we would place attributes mainly on the characters. Therefore, 4 attributes have been selected to describe and classify our characters: Family Name (only for Venetians), Doge, Churchman, King/Emperor. Those attributes were enough to cover the vast majority of our database, so almost each person had at least one relevant characteristic. As for the Family Name, we distinguished between 11 different houses, each of which had given at least one doge to the Republic during the studied period. The Doge characteristic was attributed only to people who served as doges of Venice (other Maritime Republics like Genoa had also doges in later periods). The Churchman attribute was given to individuals who held offices or prominent positions within the Roman Catholic Church, the Pope himself included. Finally, the class of Kings and Emperors is comprised solely of monarchs of foreign states, not rulers of minor fiefdoms. Near the end of the project we also added the “apostolic” class which consists of the members of families which belonged to the so-called apostolic ones and in the popular belief were the 12 families which founded the city of Venice. They used to assume important state, church, and military positions inside the Republic, while they also constituted the Minor Council of six councils to the doge. On the other hand, assigning attributes to the connections proved an impossible task mainly because of the existing time and information constraints. The only characteristic added to the connections was the Doge Succession between subsequent doges, so that a visible timeline could be followed later.
As mentioned before, the Gephi software is used to produce the related graphs. These graphs are produced by applying different filters to the database, based on the attributes introduced. Combinations and superimpositions are allowed. In Gephi, every character is represented by a “node” and every connection between the nodes by an “edge”. The complete graph is centered around the seventeen doges who ruled Venice over the period of the two whole centuries. Each doge is placed in the center with its connections forming a radial graph around it. Finally, all the smaller radial graphs are connected along the Doge Succesion line. Examples of the produced graphs for different filters and emphasis levels are illustrated below.
This project aspired to shed light on both the various internal and external connections of members of the Venetian elite. These connections include kin and friendly relations as well as political contacts and acquaintances. We set the doges of two centuries as the main characters and reference points of our research. Finding and collecting the data was a difficult and painful process. By careful examination of the database and the corresponding graphs several conclusions can be induced. For example, the relations of doges and other Venetian nobles with the outer world are numerous and denser during times of great historical events like the Crusades. This implies that Venice had an active role and was at the forefront of the developments. Additionally, the Republic retained close ties with the Church and the Papacy, probably because of geographical reasons too. However, the relations were not always amiable and depending on the interests of Venice and which family was on power (through the doge), there was friction that lead to open disagreement and excommunication a small number of times. Nevertheless, patricians used to occupy higher ecclesiastical positions within the regions controlled by Venice, frequently exercising influence in politics as well. Another point of emphasis could be the internal connections of the families, such as marriages between members of patrician families or the formation of parties with common political beliefs and goals and the subsequent political alliances or rivalries.
All the above will certainly be clearly shown and better justified when attributes to the different connections, indicating friendly or hostile attitude, are included. This and other suggestions, as the deduction of statistics and related groups from the graphs, can be subject of future work.
 Staley, Edgcumbe: The dogaressas of Venice : The wifes of the doges. London : T. W. Laurie
 Runciman, Steven (1952–54). History of the Crusades. Cambridge: Cambridge University Press
 Madden, Thomas F. (2003). Enrico Dandolo and the Rise of Venice. Baltimore: Johns Hopkins University Press. ISBN 0-8018-7317-7.