How to implement geolocation, a way to increase the relevance of search results
To provide search results customized to the user's location, add geolocation functionality. With geolocation, a query can include not only the target search terms but also latitude and longitude data pinpointing the location of the person submitting the search. Searchify can use this information to provide search results related to that location.
Why Use Geolocation?
For some types of search results, the location of the person doing the search matters. A search for "coffee shop" or "plumber" is most likely not an abstract or academic one, but a practical search for an immediate need. If you know the location from which a query is submitted, you can potentially use that information to provide more useful search results. Finding a nearby place to get coffee, or a local plumber to fix a broken water pipe, is the result most likely to make users happy.
You might want to use location data as a ranking signal for every query. Alternatively, you can let the user decide whether location is important by providing a special UI control, such as a "Find Near Me" button or a Settings checkbox, and run the location-based query only when the user selects to do so.
How It Works
Geolocation is performed in several parts:
- Index: a location (latitude and longitude) can be associated with each document in the index. The latitude and longitude are stored in the index as document variables. Location data can be associated with all or only some of the documents in the index.
- Scoring function: a built-in distance scoring function can be applied to any query to calculate the distance between the user's location and the lat/long point stored for a document.
- Query: in addition to the scoring function, pass the user's location in the query by including lat/long values as query variables. With this data, the user's location can be compared to the document location.
- User interface: You might already know the user's location: perhaps your users are authenticated and you have stored profiles for them, you are using IP location software, or your users are accessing your application through GPS-enabled devices. If not, you need to provide a way for the user to specify a location (typically using a postal code or street address). By adding a little code to convert locations into lat/long coordinates, your application is ready to put location-based result ranking into effect.
Quick Start: Copy & Paste
Tweak the following code snippets for your needs, and you're ready to go
Before you start
Add Lat/Long Data to the Documents
For those documents in your index that can be associated with a particular geographic location, store that location as a pair of document variables: one to store the latitude, and one to store the longitude.
The following code shows how to add latitude and longitude data to documents in the index, using the addDocument() method from the Java client library. The latitude and longitude values are expressed in degrees, as floating point numbers. The location used for this example is Buenos Aires, Argentina:
Define Distance Scoring Function
Searchify provides a built-in function that you can use to calculate the distance between two points. Two versions are provided, miles() and km(), for calculating distance in either miles or kilometers. For details on the function syntax, see Scoring Function Formulas.
The following example shows how to use a distance function in the definition of custom scoring function number 5. The inclusion of the negative sign will cause this function to rank search results from shortest distance to longest.
Specify Location as a Query Variable
To pass the geolocation of a particular user in a search query, use query variables. The following code shows how to include both the distance scoring function and the lat/long data as parameters to a query: