You are here: GeekEstate Blog » Development » Geeking Out On Artificial Intelligence In Real Estate

Geeking Out On Artificial Intelligence In Real Estate

As a real estate guy, sometimes I look over to my data scientist’s screen and believe what he is doing is some sort of magic — it’s like pure computer science sorcery.

I mean, they’re building a chatbot that can communicate using human language and learn from the conversations it’s having.

It’s like me every time I have a conversation with my data science team I learn something about artificial intelligence and I’d like to share that with you. So straight from the minds of 3 PhD’s in statistics, mathematics and computer science – here’s how AI is being applied to real estate.

Artificial Intelligence (AI)

Let’s start at the top, with AI. Generally speaking most true techies don’t actually call what they are doing “AI” – they break it down into a specific technology like:

  • Deep learning
  • Natural Language Processing
  • Predictive analytics
  • Machine learning
  • Supervised learning vs. unsupervised learning

AI is when a computer can do things humans need intelligence to do.

Think about all the decisions that go into assessing the value of a property – that could be one example of AI, but generally that’s a predictive analytics or deep learning model.

Predictive Analytics

Analytics play a huge role in AI. In fact, I think that analytics is truly what shapes AI. Think about the analytics from your business. Your users viewed some properties, liked a few more, opened your emails with more properties then scheduled a showing for another property- all great analytics that can be used to power AI.

Most Automated Valuation Models (AVMs) are built with predictive analytics – i.e. lots of data describing previous home sales used to predict the value of a new similar property.

The Zestimate is a good example of this, built using a Random Forests model. This model is basically the sum of a huge number of decisions made across multiple decision trees (and lots of trees make a forest).

Deep Learning

Deep learning is probably the closest thing we have to true AI. Some techies argue that deep learning is just computer scientists’ dreams and wishes but to the true nerds, deep learning is where the fun begins.

Today, my team explored a Neural Network model they built using thousands of property descriptions. A neural network is often thought of as a network of decisions made intelligently, like a human mind. But that’s just the media’s portrayal – here’s how they actually work:

In our case we took some thousands of listing descriptions and turned every word into a vector and plotted all the vectors in a huge multi-dimensional space. Next, we wanted to figure out which sentences and entire remarks were similar by examining the distances between words in a description in this multi-dimensional space.

Then we found cluster’s of words and descriptions near each other and we fed the model some words like “open concept” to find other words that were similar like “spacious” or “airy”.

To visualize, we automatically created a word cloud (shaped like our logo) that plots similar words near each other and more common words larger – check it out:

Word cloud auto-generated using an NLP Neural Network

Word cloud auto-generated using an NLP Neural Network

We even tried the word “hookup” for fun (since we are college kids) and our model returned “wine”, “bedroom” and “cellar”…which makes me wonder, what are agents writing about?

Needless to say, neural networks are fun and powerful models, but they are extremely picky and take millions of data points to become properly tuned.

Supervised and Unsupervised Learning

Just like some of us humans need supervision so to does AI. In fact, most of the workload for this type of AI is done by humans. At Structurely, our team has hand tagged tens of thousands of listing descriptions and another tens of thousands individual sentences in listing descriptions.

This data is fed into a model like TF-IDF (Term Frequency – Inverse Document Frequency) which is often thought of as “Old Faithful” in data science, to pick out and summarize listing descriptions based on the most important words  – like “open concept” or “big backyard” and forget irrelevant words overly used by agents like “beautiful”.

Unsupervised learning tells the computer to learn something on its own just from the data. It’s very similar to deep learning because both require millions of data points. Companies like X.ai who build build bots to help you schedule meetings take the unsupervised approach by feeding millions of emails about meetings to their models.

What Does All This Actually Mean?

To be honest, sometimes I have no idea. AI is becoming a buzzword in most industries and rightfully so – it’s buzz-worthy work. However, as the geeks of  real estate we need to make sure the industry is well informed of what AI actually is and what it can and cannot do.

Because AI has the opportunity to impact the most important aspect of real estate: relationships.

About Nathan Joens

Nate is the co-founder of Structurely, building artificial intelligence for real estate to help personalize interactions buyers and sellers have through messaging. Nate works with a team of 3 data scientists specializing in artificial intelligence technology who love to solve the tough problems they are faced with in real estate data.

This entry was posted in Development and tagged , . Bookmark the permalink.
  • Mike Price

    I’m fascinated by the use of big data and how it will impact the real estate industry. I just found http://www.roofai.com the other day but haven’t had a chance to dig deeper. Anyone out there know much about them?

    • I just took a look, and their pitch is really slick. It seems like much of what they do already exists, but they’re packaging everything into one bundle. I like the concept, it will be interesting to see it in action. I’ve seen some promising AI/bot models that turned out to be pretty thin in practice.

      Regarding the post, is a zestimate really classified as predictive analytics? Maybe I’m splitting hairs, but it seems to be just a comparative analytics model. It’s not “predicting” the future value of a home, it’s merely estimating the current value. Most predictive models I’ve seen have attempted to anticipate a future event.

      • Nate Joens

        Hey Sam, you’re probably correct in saying the Zestimate isn’t necessarily “predicting” something – the random forest model it uses is more or less classifying an ensemble of decisions it makes about a property compared to the current state of the market.

        However, each ‘decision’ is actually a predicted score on each ‘leaf’ of the tree – i.e. the tree asks: “is this home 3 beds – Yes/No?” then it scores it accordingly then sums the total of the leafs (decisions) based on weights that correlate to what they’ve predicted to increase or decrease the current market value of the property.

        Its kinda funny cause its making a prediction on the current market value – but I still think it does actually boil down to a prediction. Attached is a great example of a similar decision tree model

        • Great topic, either way. Lead gen platforms are going to integrate AI/predictive analytics much faster than many realize. It’s on our radar.

          • Mike Price

            Big data wise, you might want to check out what http://www.smartzip.com and the http://www.rebogateway.com products are all about. Predictive analytics being used to generate leads. I cannot speak to either as to whether or not they are effective tools but I have checked them both out. It makes sense that gathering several data points would indicate whether or not someone will list soon, divorce etc. etc. I met Avi Gupta, Smartzip’s CEO a while back. Very sharp, super nice guy that gets it.

  • “Because AI has the opportunity to impact the most important aspect of real estate: relationships.”

    I think about relationships/community constantly – not sure I’m making the jump from AI to relationships though. What specifically do you mean/think?

    • Nate Joens

      Hey Drew,

      What im thinking it boils down to is something you’ve wrote about here: http://geekestateblog.com/next-major-consumer-win-real-estate/

      when you explain: “That is…until someone changes the game, and provides a trusted environment to interact with professionals on the consumers terms.”

      Largely, the problem with agent’s lead gen strategy is lead capture forms are horrendous, and their follow up is sometimes even more miserable…I think you put it as “Like it or not, they don’t want to submit a form, talk to you for 5 minutes or trade 2 emails, only to be put on your email newsletter for eternity on some sort of crappy drip campaign”.

      So our goal at Structurely is actually to give leads a place to interact with a knowledgable AI (anonymously until they’d like more updates) on any messaging tool they prefer. This allows agents to create 1:1 conversations with leads at scale which could convert more leads by building better initial relationships (rather than a forced lead registration form)

      -> its lead generation 2.0 (using AI to reach out to your leads and help answer their questions until they are ready for your human assistance)

      The link I posted back to Structurely in this article actually borrows your 1-liner, http://www.structurely.com/leadcap2

      • For this to work well, I think AI has to get good enough that it’s not blatantly obvious the consumer is talking to a bot. I can quickly assess whether something is automated, and when I receive an automated, non personalized response…I basically discount it, and stop using service X to interact.

        I know me (and entrepreneurs in general) are not the typical consumer though.

  • Matt Patty

    Our growing Real Estate Company are big fans and supporters of SmartZip https://www.smartzip.com/ for our AI solution. Smart Zip is combining age old farming techniques, mailing out with property evaluation websites (similar to a Zestimate, we feel more accurate) and using Big Data to predict WHO to mail to, thus lowering the cost and increasing the efficiency. They are also making sure we are getting online impressions on the computers of the home owners in our target markets. Overall, like any system, it won’t work unless you put in time and effort to make it work, however we are very pleased with the systems, customer service and listings we get from our vendors at Smart Zip. Feel free to check out their web or connect with one of their knowledgeable staff or sales people.

  • Real estate industry gets lot of advantages from artificial intelligence. AI can improve communication between the real estate agent and the potential customer by ensuring that the right message is sent at the right time, to the right customer. And it helps for consumer to find properties they might be interested.

2008 - 2017 GEEK ESTATE · ALL RIGHTS RESERVED - THEME BY Virtual Results
Hosted by Caffeine Interactive