CSFinder: A Cold-Start Friend Finder in Large-Scale Social Networks

Recommending users for a new social network user to follow is a topic of interest at present. The existing approaches rely on using various types of information about the new user to determine recommended users who have similar interests to the new user. However, this presents a problem when a new user joins a social network, who is yet to have any interaction on the social network. In this paper we present a particular type of conversational recommendation approach, critiquing-based recommendation, to solve the cold start problem. We present a critiquing-based recommendation system, called CSFinder, to recommend users for a new user to follow. A traditional critiquing-based recommendation system allows a user to critique a feature of a recommended item at a time and gradually leads the user to the target recommendation. However this may require a lengthy recommendation session. CSFinder aims to reduce the session length by taking a case-based reasoning approach. It selects relevant recommendation sessions of past users that match the recommendation session of the current user to shortcut the current recommendation session. It selects relevant recommendation sessions from a case base that contains the successful recommendation sessions of past users. A past recommendation session can be selected if it contains recommended items and critiques that sufficiently overlap with the ones in the current session. Our experimental results show that CSFinder has significantly shorter sessions than the ones of an Incremental Critiquing system, which is a baseline critiquing-based recommendation system.

SOPS: A System for Efficient Processing of Spatial-Keyword Publish/Subscribe

Massive amount of data that are geo-tagged and associated with text information are being generated at an unprecedented scale. These geo-textual data cover a wide range of topics. Users are interested in receiving up-to-date geo-textual objects (e.g., geo-tagged Tweets) such that their locations meet users’ need and their texts are interesting to users. For example, a user may want to be updated with tweets near her home on the topic “dengue fever headache.” In this demonstration, we present SOPS, the Spatial-Keyword Publish/Subscribe System, that is capable of efficiently processing spatial keyword continuous queries. SOPS supports two types of queries: (1) Boolean Range Continuous (BRC) query that can be used to subscribe the geo-textual objects satisfying a boolean keyword expression and falling in a specified spatial region; (2) Temporal Spatial-Keyword Top-k Continuous (TaSK) query that continuously maintains up-to-date top-k most relevant results over a stream of geo-textual objects. SOPS enables users to formulate their queries and view the real-time results over a stream of geotextual objects by browser-based user interfaces. On the server side, we propose solutions to efficiently processing a large number of BRC queries (tens of millions) and TaSK queries over a stream of geo-textual objects.

Infrared and Raman screening of seized novel psychoactive substances: A large scale study of >200 samples

The potential of IR absorption and Raman spectroscopy for rapid identification of novel psychoactive substances (NPS) has been tested using a set of 221 unsorted seized samples suspected of containing NPS. Both IR and Raman spectra showed large variation between the different sub-classifications of NPS and smaller, but still distinguishable, differences between closely related compounds within the same class. In initial tests, screening the samples using spectral searching against a limited reference library allowed only 41% of the samples to be fully identified. The limiting factor in the identification was the large number of active compounds in the seized samples for which no reference vibrational data were available in the libraries rather than poor spectral quality. Therefore, when 33 of these compounds were independently identified by NMR and mass spectrometry and their spectra used to extend the libraries, the percentage of samples identified by IR and Raman screening alone increased to 76%, with only 7% of samples having no identifiable constituents. This study, which is the largest of its type ever carried out, therefore demonstrates that this approach of detecting non-matching samples and then identifying them using standard analytical methods has considerable potential in NPS screening since it allows rapid identification of the constituents of the majority of street quality samples. Only one complete feedback cycle was carried out in this study but there is clearly the potential to carry out continuous identification/updating when this system is used in operational settings.