Abstract:
An air quality alert system provides a signal which indicates whether or not a door or window should be (fully or partially) opened or closed. The system takes account of the air quality in a first area (e.g. outdoors) as well as the proximity of the user to a door or window so that information is only provided when the user is situated at the door or window. This avoids overloading the user with unwanted information.
Abstract:
Systems and methods to predict circadian phase of a subject use one or more sensors to track light exposure of the subject over a period and determine the current circadian phase. A prediction for future light exposure is used to predict a future circadian phase.
Abstract:
A wireless network comprises access points (APs) implementing roaming by way of handover of wireless devices between APs. Handover events over an analysis time interval are extracted from a wireless network log of the wireless network. Each handover event comprises a source AP from which a wireless device disconnects and a destination AP to which the wireless device handover events. The graph has nodes representing at least a subset of the APs of the extracted handover events and edges connecting pairs of nodes with each edge having a weight representing a count or frequency of handover events between the APs represented by the pair of nodes connected by the edge. A network diagnostic or visualization task is performed using the HF network topology graph.
Abstract:
A data base system is disclosed for storing meta-data about labeled predominantly non-text items in a collection of labeled predominantly non-text items (lpnt items). The system comprises a storage facility (10), a keyword generation facility (20) and a meta-data enrichment facility (30). The storage facility (10) stores respective meta-data for lpnt items (i) within said collection, said meta-data comprising at least one lpnt item classification (gi) and a 1pnt item description (Ti), said classification defining classes. The keyword generation facility (20), which generates keyword data for respective classes, comprises a retrieval module for retrieving terms from lpnt item descriptions of 1pnt items associated with said classes. The keyword generation facility (20) further comprises a term relevance determining module for assigning a relevance of said terms for said classes, for selecting terms from said retrieved terms on the basis of their assigned relevance and providing these selected terms as candidate keywords. The system further comprises a metadata enrichment facility (30) for adding a subset (Ki) of the provided candidate keywords (K(c)) as keywords to the meta-data of lpnt items of a class.