# Feature overview The ranking boost and suppression feature for recommended items lets you manually adjust specific items' positions in recommendation results based on user location (LBS) and item distance dimensions. It offers a direct and effective way to influence recommendations, helping you flexibly achieve business goals such as promoting new products, clearing inventory, suppressing negative items, or optimizing localized recommendations. This guide explains how to configure item ranking boost and demotion using the console. ## Key features * Boosting configuration: increase the recommendation weight of specified items to make them more likely to appear in results. * Demotion configuration: decrease the recommendation weight of specified items to make them less likely to appear, with support for triggering suppression based on distance dimensions. * Weight factor: Supports configuring an integer weight factor from \-100% to 100%. The larger the value, the more significant the effect of boosting or demoting. * Distance dimension support: Allows configuring weights based on user geographic location (LBS) and actual item distance range to achieve regional boosting or demoting control. For example, items exceeding the user\-set distance threshold automatically apply demotion. ## Typical use cases * This strategy targets items located outside a defined geographical range from the user (e.g., items in cross\-regional warehouses or products with delivery restrictions). By **"Demotion" (applying a negative weight)** to these items, you can effectively reduce their exposure in the recommended results. # Prerequisites To use location capabilities, ensure the following prerequisites are completed: 1. The "item dataset" used in the recommendation scenario must have "longitude" and "latitude" fields configured.

# Configuration steps ## Add a new geographic location rule :::warning \*Note The system does not support priority configuration. When an item matches multiple configurations, the platform combines the ranking weight increases and decreases from all configurations to calculate the final weight factor. For example, if item A matches a "boost 50%" rule and a "demote 30%" rule, the platform will ultimately increase the item's weight by 20%. ::: Follow the steps below to boost or reduce the ranking of specific items: 1. Go to **“Experience Center” — “Recommendation Experience” — “Strategy”** and click “Boost and Demotion” to enter the ranking adjustment configuration area.

2. After entering the rule configuration page, please fill in the following information:

|**Configuration item** |**Description** | |---|---| |Rule name |Set a simple and clear name, for example: "Product demotion beyond 20 km in the Beijing area". | |Scope |1. Navigate to "calculated fields" and select "distance between user and item".|\ | ||\ | |![图片](https://p9-arcosite.byteimg.com/obj/tos-cn-i-goo7wpa0wc/c8a91c86f6ec48229b9b965411385a71 =400x) |\ | ||\ | |2. In the condition settings, select "greater than" or "less than" from the dropdown, enter the distance value (e.g., 20) in the input box, and choose the unit (e.g., km).|\ | ||\ | |![图片](https://p9-arcosite.byteimg.com/obj/tos-cn-i-goo7wpa0wc/f81638eabfc94b3c8a239d7278c5e1b0 =396x) |\ | |:::tip \*What if an item has multiple coordinates?|\ | |If an item has multiple sets of longitude and latitude coordinates, the platform will automatically select the set closest to the user when calculating the distance.|\ | ||\ | |:::| |Boost / Demotion configuration |Support configuring integer weight factors from \-100% to 100%. The larger the value, the more significant the boost or demotion effect. | 3. Click “OK” to create a new rule on the platform. The new boost or demotion ranking rule will take effect in the experience interface on the next recommendation request. It is recommended to continue reading the [Experience geographic location rules](/docs/85296/1963520#b8ff614f)section to verify the configuration effect. 4. After verifying the effect on the experience center matches expectations, click “Sync Changes to API.” The online API interface will also apply the new boost or demotion ranking rule. 5. The `context` object in the API request must include the `location` field. The format example is as follows. For details, please see [Context](/docs/393/92213#65900099). ```JSON {"context":{"location": {"longitude": 116.4, "latitude": 39.9}}} ``` ## Experience location rules After configuring the promotion/demotion rules related to location, verify their effects on the recommendation experience interface to observe recommendation adjustments based on user location (LBS) and item distance. :::tip \*Prerequisites You must have completed the "Add a new geographic location rule" steps above. ::: 1. **Access the recommendation experience interface** Navigate to "Experience Center" \> "Recommendation Experience" in the console and select the target recommendation configuration (e.g., homepage recommendation, product list page). 2. **Simulate user location and item distance** In the right "Test User" area, simulate scenarios of user location boost or demotion by specifying location capabilities:

|**capability** |**Capability description** | |---|---| |Get my location |Obtain your current location to simulate a test user; this requires you to grant location permission to the browser.|\ | |:::tip \*Description|\ | |If the platform does not automatically obtain coordinates after clicking "Get my location," please check the following:|\ | ||\ | |1. Is the location capability enabled on the computer?|\ | |2. Does the computer or browser restrict the browser’s access to location data?|\ | ||\ | |If browser location access was previously denied, restart the computer to reset permission restrictions.|\ | ||\ | |:::| |Specify user location |Manually enter latitude and longitude coordinates (e.g., `longitude: 39.9°`, `latitude: 116.4°`) | 3. **Review recommendation changes** Click "Recommend Again". The system applies the configured boost/demotion rules using the simulated user location and actual item distance (requires item dataset with configured latitude and longitude). For example: * **Near\-distance item boost effect**: if the rule sets "boost items within 5 km of the user by 50%", those items rank higher in the recommendation list. * **Long\-distance item demotion effect**: If the rule sets "demote items beyond 20 km from the user by 80%", items exceeding this distance will be ranked lower (or receive less exposure) in the recommendation list. --- # Frequently Asked Questions **Q: Why do the recommendation results in the preview differ from those online?** A: The preview recommendation results are for demonstration only; actual online users will trigger rules based on their real geographic location and item distance. **Q: Why don’t I notice any difference after configuring the rules?** A: If you do not observe the expected effect, please check: 1. Whether the item dataset correctly configures the "longitude" and "latitude" fields; 2. Whether the "distance configuration" in the rules matches the test scenario; 3. Whether the weighting factor is set appropriately (the larger the value, the more noticeable the effect). **Q: Why don’t I have the "user\-item distance" option?** Answer: Please check the item dataset to see if the fields are configured with the attributes "longitude" and "latitude". If you need to update the field attribute configuration, please contact the VolcEngine technical team. --- # Related documents 1. For other boost and demotion capabilities, refer to [Item Attribute Boost and Demotion](/docs/85296/1925098) 2. If you want to boost or demotion items based on "Launch time" or "new arrival time," please refer to [ ](/docs/85296/2096788)

Viking AI Search