Content request identification via a computer network

Systems and methods of identifying information resources for content item placement via a computer network are provided. A data processing system can identify clusters associated with an information resource, and determine a weight of the plurality of clusters for the information resource. An association metric for the information resource can be determined, and the information resource can be identified as a specialized information resource based on the association metric. The data processing system can select a placement criterion for a content item based on information received from a content provider computing device via the computer network, or information received from a network session between at least one end user computing device and at least one content publisher computing device via the computer network. Based on the placement criterion, the specialized information resource can be selected as a candidate for placement of the content item.