Valuation Workflows Within ArcGIS

welcome to this envisioning Center presentation on valuation workflows within within ArcGIS my name is Tim fella I'm the Global Business Development Manager for Lane administration and I'm joined here by my close colleague Katherine Smyth who's a solution engineer supporting our international land administration business so around the world property Assessors face a variety of challenges these include maintaining the trust and confidence of the public deriving fair and equitable values minimizing appeals discovering untaxed property whereas ArcGIS has long been used to efficiently and accurately maintain land records and their associated geometries but the emergence of web GIS and the advancement of spatial modeling and statistical tools within arcgis pro we are seeing opportunities to leverage arcgis to support a growing number of property evaluation workflows most of the people in this room are probably familiar with the golden rule in real estate which is location location location and the importance it has on influencing it either the increase or decrease in property values but given the importance of location it's only natural that GIS be leveraged within the property evaluation process and there are a variety of ways it can be some of those that you see here include you know collecting or updating data in the field - geo enrichment property attributes using distance to certain features or perhaps taking data from the living atlas - identifying trends and patterns within your data to spatial modeling for mass valuation the collection evaluation capabilities within ArcGIS can be organized and grouped according to these four areas which we're going to discuss today the first of which is field operations and this includes managing your field workforce collecting property assessment data in the field the second one being value analysis which focuses on data exploration and value prediction the third area being appeals defense covers explaining your valuation process and methodology and simplifying the way in which appeals can be submitted and then lastly public data access and this includes helping to prevent Appeals but also lend itself to improving transparency and limiting front office traffic for common questions that the public may have today what we're gonna do is we're gonna walk you through a fictitious reassessment process and highlight just a few examples of workflows that can be supported through the application of GIS for those of you who may be new to what a property assessment looks like a reassessment it's really the process of revaluing all the land in an area using statistical models that take into account property characteristics such as size condition and equitably adding new values across an entire geographic area traditionally these models have used only characteristics about the properties concerned but it has been found that values can be made even more equitable by taking spatial characteristics such as distance to public transportation or other landmarks meaningful to people living in that area in addition there are other population-based characteristics that can be added using the collection of global authoritative data within the living Atlas so at this point what we're seeing here is just a fictitious kind of public notice that someone may receive about a reassessment process taking place and the link to where I can get more information on the process itself so what we're gonna do is click on that link and now I'm putting on my my public taxpayer hat on and what we're at now is a an office of tax and revenue public facing website that's been built on ArcGIS hub so coming to this website you can see if there's a lot of information about when the reassessments taking place why a reassessment is occurring what the objectives of it are the process that's going to be undertaken there's also a variety of applications and data that can be down loaded applications where you can query information about your particular parcel the characteristics about it perhaps which value valuation district you fall within comparable sales and whatnot it's a lot of data here but before I dive into looking at some of these applications what I'm going to do is turn it over to my colleague Catherine to talk about the data collection process thanks Tim as a field appraiser I've been sent out into the field to gather more property characteristics about a new development in an area that would well be reassessing in order to gather these property characteristics I'm using survey one two three I'll go ahead and open my property survey and start collecting I can see that I can add the parcel ID as well as the owner and my name as well as some other characteristics that have been configured specifically by my organization I'm collecting information about a residential property so I'm given a bunch of parameters that I can choose from this particular property is excellent it has four rooms and three bathrooms I know that the lot size is 2,300 square feet based on the legal document when I'm done I can even take a picture if I want to and this picture will be ascribed to the data itself after I'm all finished and I know that I've collected everything that I need to collect I'll go ahead and submit that survey once the survey has been submitted the property characteristics for that specific property will be updated and the data can then be shared on our public facing application so the taxpayers like Tim understand the different property characteristics that were interested in and that we're using specifically for valuation Tim thank you Catherine so as I mentioned before there's a lot of data in applications that have been put out by the office attacks in revenue to help inform the public and try to address front office traffic for common questions that may arise I'm gonna scroll down here to this particular application attacks parcel viewer because I'm interested in understanding the data that's been collected as part of the reassessment process I understand also a type of characteristics that they have documented so if I go here to the tax parcel viewer I can simply type in my address and query the information about my property so here it brings me down to the block I live on and I can select the property as you can see here pull up information about the the owner address assessment value number of bedrooms and rooms so it answers a lot of common questions that maybe someone in the public may be wanting to inquire about as you see here there's a couple other applications that have been published including this valuation districts and comparable sales finder but before I open those up I'm going to turn it over to Katherine to talk about some of the analyses and processes that go into creating these thanks Tim I'm using ArcGIS Pro to do more analysis on data within my area I'm specifically making use of something called tasks tasks can be configured to step through a workflow so that you can recreate it time after time in this case I'm going to explore my data and then I'll be enriching it with some other property characteristics as well as running a spatial model but let's look at what that data exploration process looks like first in order to start out I'll open a geoprocessing tool called exploratory regression the exploratory regression tool allows me to enter a dependent variable in this case I'm interested in seeing sales price and using candidate explanatory variables to see which of these variables impacts the sales price in my area so as you can see I've chosen the usual suspects bathroom he rooms bedroom as well as stories styles structure and condition right these are common characteristics that I have some idea of how they may influence my price but in order to see more information about this I can run the exploratory regression tool I also have a number of search criteria that I can set that can be based off of the standards that might organization sets out for me once I run the exploratory regression tool I'm given a number of models that have passing values based on the criteria that I've run and I can see the different variables that are part of that passing model so I can see that land areas significant improved property understanding bathrooms etc as well as heat and surprisingly fireplaces because this is on the East Coast so it's cold there right so that's interesting it allows you to see different characteristics and how they may influence the sales price in a region just understanding that can have an impact on how I run further tools the next step that I want to perform is to understand different neighborhoods for valuation in my district so I'm going to use the spatially constrained multivariate clustering tool this tool takes into account a number of parameters that allow different neighborhoods to be formed based on how I tell the tool to work so here I have the input features as my properties which you can see on the map as well as output features which will just be the output of the clustering I can assign the analysis fields here and in this case I would assume that the same the same variables that had statistical relevance in the exploratory regression tool would also be characteristics that I would want to put in in terms of identifying my neighborhoods for valuation once that tool is run then I'm given an output that shows all of the different neighborhoods within my area in order to make this more accessible to the public I'm going to draw a polygon around it providing this information to taxpayers like Tim allows them to have more heads-up perspective on where they're living and what characteristics may influence their property value great so here we are again back on the office of tax and revenue website and I'm querying information about my property and how it relates the reassessment process so I've already gone through the tax parcel viewer and is able to look up the characteristics of my own property and those around it but I can also open up this valuation district and similar as the parcel viewer we can put in the address and zoom in to the valuation district that I fall within so I might just be useful information to have to understand which properties are similar to my own and what valuation district I fall within but maybe I want to understand what are some of the comparable sales within my locality and what the characteristics are about those as well so I can go to this residential or comparable sales finder application it has a few functions in it one of which is you can query by the radius or distance from your property in which you want that search to take place you can also set the number of properties or comparable sales you want to pull for you so in this case I'm gonna set it for one kilometer and for six different properties that I want to be able to find comparable sales for so again I'm gonna plug in my address and here it is pulling up what we see here is six different properties within close proximity of where sales have occurred and it's similar to the other applications what you can do is understand the information about those properties this point I also see that there's another application down below that's been embedded in the website that's focused on property tax assessment appeal but before I talk about a property tax assessment appeal or plugin information I'd turn it back to Kathryn to talk about the process for predicting values sure thanks so back with my GIS analyst hat on I've been tasked to look at a specific neighborhood and to maybe understand the different drivers within that neighborhood that may be affecting value that we haven't taken into account yet in this case I've chosen neighborhood 39 and what I'm going to do is enrich the data using the geoprocessing enriched tool this just takes an input feature and also produce as an output but what's cool about it is it allows me to put multiple variables depending on what country I'm in and depending on the information available in here to understand more of the profile of the people living in this neighborhood in this case I'm interested in just visualizing the per capita income of my neighborhood to see if it may be statistical a statistically relevant when I perform evaluation modeling so I'll click OK and when I run that the outcome here allows me to visualize income based on location as we can see it's not surprising right that higher income properties are closer to the center of the city while lower income are on the outside I'm also interested in the impact that different spatial locations may have on my property value so I'm also going to prepare my data by running the NIR tool the NIR tool calculates the distance between a specific feature no matter what it is and my parcel data if that's the input here I put the input features as my personal data as well as added the NIR features as metro stations around DC because I know from living there that it's important at least to me to live in it next to a metro stop so that I can get around but whether this will influence my property valuation or not remains to be seen it's with this information that we can gain a better understanding of all of the different characteristics that go into a property valuation and by making this information available to the public itself then we can allow tax payers like Tim to either more informed do a more informed appeal or move forward from there so here we are going back to the office of tax and revenue website and scroll down here to the property tax assessment Pio application the intent of this is to be able to simplify the process of submitting Appeals for the valuations I've been derived so what you can see here is again plugging in your address for where you're located or where your property is located you can also plug in information about whether you're the owner you know the parcel identification number information about the property itself and for instance reasons for filing the appeal as well as your contact information so all the information would be commonly requested as part of the appeal process but what I'm going to do is again turn it back to Katherine to dive into some further details on the valuation process itself thanks Tim now as an Assessor valuer for my property I have a number of models that I may run on a given area in this case I'm going to run a spatial model called forest based classification and regression forest based classification and regression is based on the idea that I have many decision trees that will help me visualize whether a certain characteristic is important to sale price or not right so in the input I'm going to put my neighborhood I'd also like to mention on a side note that this this tool in the background is running a machine learning process so we're going to run the tool iteratively to train the model so that eventually we will be able to predict sale price in this area which will allow me to fill in missing missing data that I may have or data that isn't as as up-to-date so in this case I have sale price for about twenty percent of the data in this specific region so I'm going to input my region neighborhood 39 and input sales price as my variable to predict and then I'll add all of the explanatory training variables such as my usual my usual suspects are all there right but I'm also going to incorporate the other augmented training variables as well that we identified in the last one so that means per capita income as well as this explanatory distance feature which is two metro stations right so I can add as many different characteristics here as I would like which allows me to in further on an acceptable model this tool provides many different outputs but because we're training once and then we're going to be running the tool iteratively I'll just show you the results from the first from the first go-around I think that one of the most valuable outputs of this tool is the variable importance table here I've standardized the variable importance by z-score and you can see these are the names of all of the variables that I've entered so this is showing the relationship that the tool is made between these different variables as well as the sale price okay and at the end I can see that my model has a little bit of fine-tuning that needs to happen right so I would fine-tune this model over and over again until it comes out with a final result that I'm comfortable with once I reach that result then I can visualize that information here so what you're looking at at the moment is the result of a prediction of sale price that's taking into account that 20% of sales that I had before and then using the relationship between all of the variables to help predict the other 80% I can later run this through any of my models that I already have unlike coefficient of dispersion to understand whether it's accurate enough for my needs but in this case I've just done a simple ratio study by comparing it to sales ratio in the area so I can visually just look at the different patterns that are happening so right now I'm seeing that I have a higher sales price prediction closer into the city as one might expect but when I'm visualizing my sales ratio continuously I'm seeing that there's a higher discrepancy out towards the outer edge of this neighborhood so that may help me identify vertical inequity another way to visualize this information is through manual intervals so the white areas are areas that fall within the threshold that I'm comfortable with if i zoom in further than I'm able to see the outliers in red and dark blue that I may need to take another look at that may be telling me that my data isn't clean enough for example if I missed a condo and there are many stories to the building and maybe it got collapsed or if these properties are not what I think they are it could also be saying that my evaluation is off so all of this is just a visualization to help me understand my data more as the model gets stronger and stronger I can test that confidence level as I go and that will allow me to have a more impactful sales sales forecast for example so if I had data from ten years ago I could iteratively run these models just to test out and see what the outcome was and whether that fit into my actual workflow effectively or not when I was doing this this specific study I also realized that I hadn't taken into account another really important geographic feature which are parks and if you look back at the sales ratio as well as the predictions you can see that generally trending property prices are projected to be higher around parks so that may be something else that I want to go back and add as I continue to fit this model all of this information can further inform my my personal valuation of this area and make sure that tax payers like Tim have an equitable valuation of their property and meanwhile have access to that hub site so that they can see what I'm doing throughout the entire process and be more familiar with what I so hopefully this presentation has shed some light on the variety of ways in which GIS can support many common property valuation workflows ranging from the field data to collection process to spatial modeling as well as communicating with the public about the reassessment process and the data collected on their property by leveraging these tools within ArcGIS we were hoping valuation offices can help reduce costs reduce a number of appeals that are coming in increase public engagement as well as the transparency and trust that the public has in their evaluations and obviously increase the efficiency in their operations as well I want to thank you for joining us today for the session focused on valuation workflows with an ArcGIS if you're interested in learning more about what we showed we inter encourage you to attend some of the related sessions throughout the next couple of days so you can see here there's a variety of sessions related to land administration these include both property valuation as well as parcel management happy to answer any questions we also have a survey that we would love for you to fill out afterwards and once again thanks a lot for sitting through this session you you

Channel: Esri Industries

Views: 1,317

Rating: 4.7647057 out of 5

Keywords: Esri, ArcGIS, GIS, Geographic Information System, ArcGIS Pro, Valuation, Land Records, Real Property, Home Values, Property Values, Spatial Analysis, Real Estate, Assessor, Appraiser, Assessing, Land Administration

Id: ico5G1hJ8Yk

Length: 24min 7sec (1447 seconds)

Published: Mon Aug 19 2019