Location Prediction for Tweets.

Geographic information provides an important insight into many data mining and social media systems. However, users are reluctant to provide such information due to various concerns, such as inconvenience, privacy, etc. In this paper, we aim to develop a deep learning based solution to predict geographic information for tweets. The current approaches bear two major limitations, including (a) hard to model the long term information and (b) hard to explain to the end users what the model learns. To address these issues, our proposed model embraces three key ideas. First, we introduce a multi-head self-attention model for text representation. Second, to further improve the result on informal language, we treat subword as a feature in our model. Lastly, the model is trained jointly with the city and country to incorporate the information coming from different labels. The experiment performed on W-NUT 2016 Geo-tagging shared task shows our proposed model is competitive with the state-of-the-art systems when using accuracy measurement, and in the meanwhile, leading to a better distance measure over the existing approaches.

Manipulation of micro-particles through optical interference patterns generated by integrated photonic devices.

Micro-particle transport and switch governed by guided-wave optical interference are presented. The optical interference, occurring in a directional coupler and a multi-mode interferometer made by inverted rib waveguides, results in a specific evanescent field dependent on wavelength. Through a detailed theoretical analysis, the field of induced optical force shows a correlative pattern associated with the evanescent field. Experimental results demonstrate that 10 µm polystyrene beads are propelled with a trajectory subject to the interference pattern accordingly. By launching different wavelengths, the polystyrene beads can be delivered to different output waveguide ports. Massive micro-particle manipulation is applicable.

Study of disklike microdroplet cavities directly coupled to liquid-core waveguides.

A design of microfluidic devices is presented to integrate single-mode, liquid-core waveguides with microfluidic channels that generate and deliver disklike emulsion microdroplet cavities doped with an organic dye. The microcavity modes can be directly coupled to the liquid waveguide. Cavity-enhanced spontaneous emission was observed at the waveguide with low pump pulse energy.