I have a software request, if anyone has any time.
Years ago, I was searching a seemingly dead end genealogy line. This one was important to me, so I was giving it a little extra effort. Ancestral File, pension records, discovered state records, census information, etc. did not provide any tangible information. I went to all the major genealogy sites, searched newsgroups, etc. but didn't find any new information. I had exhausted the traditional search venues and was resigned to random searches. This search took place while I was finishing my MS in CS at BYU. I was still primed from finals week, but wanted to do something else. ;-) Anyway, I found family names in books of state records. Again finding my all too familiar dead end, I just started looking at books that were sitting adjacent on the shelf. I saw my family names listed 'next to' other's names. When those names were not in alphabetical order, I seemed to pay more attention. Eventually, those names started to seem familiar. On the original census reports (not the alphabetical ones), those same names appeared. Sometimes, they were listed before and other times they were listed after, but all the names seemed to be part of a group. In one case in about 1800, there was a migration from SC to TN. Several of those families were listed together again. Then several of them moved on to AL and eventually to the Republic of Texas. This was a pretty interesting find as it showed that these people, who were completely disjoint on the alphabetical census and in Ancestral File, seemed to know each other and / or be related. I made mental and some written notes of all of this and filed it away. Return trips would find additional 'relationships' under corollary circumstances. Still, it could not get me to the next generation. Years later, I was in the SLC genealogy library. Ancestral File is the same there as anywhere, so I started looking around again. This library was 'cool', because it actually had sources that were simply referenced elsewhere. I started looking at sources and came across journals and family histories from some of these other families. As it turns out, some of these families and mine were related by marriage. When I read their histories, I found references to my family members and new details about them. I started to put together names and relationships that had previously eluded me. This bigger picture helped me start to see who was a part of my family and who simply shared the same last name. I found a couple of histories and journals with these types of references - keep in mind that the sources / subjects were not in my direct family tree, but rather part of the same community. Still, without that, I would not have found out so many details about my own line. In one case, I even found a general store receipt listing what my g-g-g-g-grandfather purchased in the early 1800's. Okay, so back to software. As we put more and more databases online, we gain tremendously, but we tend to lose some of the detail. Also, we are all doing the same things - searching names and references. My discoveries were found, not using the typical computer methods, but rather techniques very familiar to the experienced genealogy researcher. I would like to develop some software that would use data mining techniques to help me answer the following questions (and more): Who might a specified individual have known? Who lived near them? Do my search subjects and their 'neighbors' share any common relatives? Did they migrate? If so, with whom? Did they have business dealings? If so, with whom? Do any of these neighbors and associates have family histories? Do those neighbors' and associates' histories reference my family? Do those neighbors' and associates' histories reference names / places that reference my family? There are probably more questions that fit with the direction of this list, but you get the idea. Again, these questions are not new to the artistic genealogy researcher, but don't seem to have yet made it into the mainstream computer software (that I have seen, anyway). The method that I am suggesting is more of a fine tuning type of search that applies to some searches and not to others. To pioneer and small town histories, it absolutely applies. However, to Paris in 2007, you would be hard pressed to make any sense (without further refinement) of the plethora of data. With the new FamilySearch API, some of these searches may now be possible. Others will follow as new reference material is placed on line and indexed. At that point, they can also be linked in. Genealogy will always be an art form, but building this type of software will definitely help lesser experienced researchers make new discoveries. So, what do you think? Any one interested? Ideas? There is probably a good Master's Thesis in here. (hint, hint). Thanks, Steve ---------------------------------------------------------------------------- http://www.mccownclan.com <BLOCKED::http://www.mccownclan.com> "Chance favours the prepared mind." -- Louis Pasteur ----------------------------------------------------------------------------
_______________________________________________ Ldsoss mailing list Ldsoss@lists.ldsoss.org http://lists.ldsoss.org/mailman/listinfo/ldsoss
