On top of an enterprise social platform, we are building a smart social QA system that automatically routes questions to suitable employees who are willing, able, and ready to provide answers. Due to a lack of social QA history (training data) to start with, in this paper, we present an optimization-based approach that recommends both top-matched active (seed) and inactive (prospect) answerers for a given question. Our approach includes three parts. First, it uses a predictive model to find top-ranked seed answerers by their fitness, including their ability and willingness, to answer a question. Second, it uses distance metric learning to discover prospects most similar to the seeds identified in the first step. Third, it uses a constraint-based approach to balance the selection of both seeds and prospects identified in the first two steps. As a result, not only does our solution route questions to top-matched active users, but it also engages inactive users to grow the pool of answerers. Our real-world experiments that routed 114 questions to 684 people identified from 400,000+ employees included 641 prospects (93.7%) and achieved about 70% answering rate with 83% of answers received a lot/full confidence.

Various methods have been proposed to help find answerers for a question in QA communities, but almost all work heavily depends on users' QA history. In this paper we seek to investigate the feasibility of leveraging users' non-QA social activities as a way of gaining insight into their question answering behavior. We collected data of 4,484 users on a QA community in an enterprise and examined the relationship of their QA behaviors and non-QA activities supported by other social tools. We found that the two sets of behavioral indicators are significantly correlated. The top user group for non-QA activities outperformed lower groups in both number and quality of the answers. Regression analysis showed involvement in personal blog, micro-blog and forum and replying and recommending behaviors in non-QA communities were predictive of a user's likelihood of answering questions. Online observations provided a qualitative understanding. Design implications and future work were discussed.

This paper proposes an efficient layer-correlated scheme to educe the bit cost for motion vectors in the 3D wavelet coding, previous works show that incorporating motion alignment into lifting structure enables the 3D wavelet coding to provide a competitive performance to the state-of-the-art JVT standard. In general, the temporal wavelet decomposition consists of multiple layers, while each layer adopts one set of motion vectors to achieve high coding efficiency and temporal scalability. Since the current schemes code these MVs independently, this greatly increases the bit cost for coding MV. In order to reduce the motion cost, the proposed scheme performs motion estimation considering the MV correlation among layers. Several modes are proposed to describe the different local correlations at the macroblock level. By an R-D optimized mode selection engine, the proposed scheme can save up to 33% bits of MVs at the similar texture quality.

Using an array of photo shots, the concentric mosaic offers a quick way to capture and model a realistic three-dimensional (3-D) environment. We compress the concentric mosaic image array with a 3-D wavelet transform and coding scheme. Our compression algorithm and bitstream syntax are designed to ensure that a local view rendering of the environment requires only a partial bitstream, thereby eliminating the need to decompress the entire compressed bitstream before rendering. By exploiting the ladder-like structure of the wavelet lifting scheme, the progressive inverse wavelet synthesis (PIWS) algorithm is proposed to maximally reduce the computational cost of selective data accesses on such wavelet compressed datasets. Experimental results show that the 3-D wavelet coder achieves high-compression performance. With the PIWS algorithm, a 3-D environment can be rendered in real time from a compressed dataset.

A motion compensated lifting (MCLIFT) framework is proposed for the 3D wavelet video coder. By using bi-directional motion compensation in each lifting step of the temporal direction, the video frames are effectively de-correlated. With proper entropy coding and bitstream packaging schemes, the MCLIFT wavelet video coder can be scalable in frame rate and quality level. Experimental results show that the MCLIFT video coder outperforms the 3D wavelet video coder with the same entropy coding scheme by an average of 1.1-1.6dB, and outperforms MPEG-4 coder by an average of 0.9-1.4dB.