Developing a forest data portal to support multi-scale decision making

Forests play a pivotal role in timber production, maintenance and development of biodiversity and in carbon sequestration and storage in the context of the Kyoto Protocol. Policy makers and forest experts therefore require reliable information on forest extent, type and change for management, planning and modeling purposes. It is becoming increasingly clear that such forest information is frequently inconsistent and unharmonised between countries and continents. This research paper presents a forest information portal that has been developed in line with the GEOSS and INSPIRE frameworks. The web portal provides access to forest resources data at a variety of spatial scales, from global through to regional and local, as well as providing analytical capabilities for monitoring and validating forest change. The system also allows for the utilisation of forest data and processing services within other thematic areas. The web portal has been developed using open standards to facilitate accessibility, interoperability and data transfer.

Learning sentiment classification model from labeled features

We propose a novel framework where an initial classifier is learned by incorporating prior information extracted from an existing sentiment lexicon. Preferences on expectations of sentiment labels of those lexicon words are expressed using generalized expectation criteria. Documents classified with high confidence are then used as pseudo-labeled examples for automatical domain-specific feature acquisition. The word-class distributions of such self-learned features are estimated from the pseudo-labeled examples and are used to train another classifier by constraining the model's predictions on unlabeled instances. Experiments on both the movie review data and the multi-domain sentiment dataset show that our approach attains comparable or better performance than exiting weakly-supervised sentiment classification methods despite using no labeled documents.

A novel framework of training hidden Markov support vector machines from lightly-annotated data

Natural language understanding (NLU) aims to map sentences to their semantic mean representations. Statistical approaches to NLU normally require fully-annotated training data where each sentence is paired with its word-level semantic annotations. In this paper, we propose a novel learning framework which trains the Hidden Markov Support Vector Machines (HM-SVMs) without the use of expensive fully-annotated data. In particular, our learning approach takes as input a training set of sentences labeled with abstract semantic annotations encoding underlying embedded structural relations and automatically induces derivation rules that map sentences to their semantic meaning representations. The proposed approach has been tested on the DARPA Communicator Data and achieved 93.18% in F-measure, which outperforms the previously proposed approaches of training the hidden vector state model or conditional random fields from unaligned data, with a relative error reduction rate of 43.3% and 10.6% being achieved.

Obstacles and solutions to maximising urban biodiversity throughout the lifecycles of major developments in England

Thus far, achieving net biodiversity gains through major urban developments has been neither common nor straightforward - despite the presence of incentives, regulatory contexts, and ubiquitous practical guidance tools. A diverse set of obstructions, occurring within different spatial, temporal and actor hierarchies, are experienced by practitioners and render the realisation of maximised biodiversity, a rarity. This research aims to illuminate why this is so, and what needs to be changed to rectify the situation. To determine meaningful findings and conclusions, capable of assisting applied contexts and accommodating a diverse range of influences, a ‘systems approach’ was adopted. This approach led to the use of a multi-strategy research methodology, to identify the key obstructions and solutions to protecting and enhancing biodiversity - incorporating the following methods: action research, a questionnaire to local government ecologists, interviews and personal communications with leading players, and literature reviews. Nevertheless, ‘case studies’ are the predominant research method, the focus being a ‘nested’ case study looking at strategic issues of the largest regeneration area in Europe ‘the Thames Gateway’, and the largest individual mixeduse mega-development in the UK (at the time of planning consent) ‘Eastern Quarry 2’ - set within the Gateway. A further key case study, focussing on the Central Riverside development in Sheffield, identifies the merits of competition and partnership. The nested cases, theories and findings show that the strategic scale - generally relating to governance and prioritisation - impacts heavily upon individual development sites. It also enables the identification of various processes, mechanisms and issues at play on the individual development sites, which primarily relate to project management, planning processes, skills and transdisciplinary working, innovative urban biodiversity design capabilities, incentives, organisational cultures, and socio-ecological resilience. From these findings a way forward is mapped, spanning aspects from strategic governance to detailed project management.

Joint sentiment/topic model for sentiment analysis

Sentiment analysis or opinion mining aims to use automated tools to detect subjective information such as opinions, attitudes, and feelings expressed in text. This paper proposes a novel probabilistic modeling framework based on Latent Dirichlet Allocation (LDA), called joint sentiment/topic model (JST), which detects sentiment and topic simultaneously from text. Unlike other machine learning approaches to sentiment classification which often require labeled corpora for classifier training, the proposed JST model is fully unsupervised. The model has been evaluated on the movie review dataset to classify the review sentiment polarity and minimum prior information have also been explored to further improve the sentiment classification accuracy. Preliminary experiments have shown promising results achieved by JST.

Efforts towards adaptability enhancement:how manufacturing and service industries can learn from each other

This paper discusses the recent emerging efforts for adaptability enhancement of Japanese industries to cope with a volatile demand environment. It is based on an analysis of data obtained from respondent companies. The analysis is focused on both manufacturing and manufacturing- related service industries such as construction/maintenance, software supply, manufacturing consultation and logistics industries to highlight their current situation, the sense of crisis in Japanese companies and possible future directions in relation to the two industry sectors. The principal conclusion is that for most companies consideration of a revision or modification to its cost structure is an essential requirement for survival in the global competitive environment.

Originator or propagator? Incorporating social role theory into topic models for twitter content analysis

A large number of studies have been devoted to modeling the contents and interactions between users on Twitter. In this paper, we propose a method inspired from Social Role Theory (SRT), which assumes that a user behaves differently in different roles in the generation process of Twitter content. We consider the two most distinctive social roles on Twitter: originator and propagator, who respectively posts original messages and retweets or forwards the messages from others. In addition, we also consider role-specific social interactions, especially implicit interactions between users who share some common interests. All the above elements are integrated into a novel regularized topic model. We evaluate the proposed method on real Twitter data. The results show that our method is more effective than the existing ones which do not distinguish social roles. Copyright 2013 ACM.

Innovation across the food chain – the Hungarian case

The small and medium sized enterprises (SMEs) in the Hungarian agri-food sector play determining role. The innovation capacity (efforts, activities and results) however of the individual SMEs is very limited. Food production (including SMEs) has to fulfil food safety requirements in a rapidly increasing extent, which implies a continuous innovation and development process from all market players. In Hungary the agri-food chain had to face a suddenly increased competition especially after the EU enlargement. Based on survey data this paper examines the efforts, activities and results in knowledge acquisition, utilisation, coordination and transfer in the Central Hungarian food SMEs. We have found (using ordered logistic regression) that R&D expenditures, achieved innovations, export/import orientation as well as the networking activity of the SMEs play significant role in market development.

Hedging under liquidity constraints

Although risk management can be justified by financial distress, the theoretical models usually contain hedging instruments free of funding risk. In practice, management of the counterparty risk in derivative transactions is of enhanced importance, consequently not only is trading on exchanges subject to the presence of a margin account, but also in bilateral (OTC) agreements parties will require margins or collateral from their partners in order to hedge the mark-tomarket loss of the transaction. The aim of this paper is to present and compare two models where the financing need of the hedging instrument also appears, influencing the hedging strategy and the optimal hedging ratio. Both models contain the same source of risk and optimisation criterion, but the liquidity risk is modelled in different ways. In the first model, there is no additional financing resource that can be used to finance the margin account in case of a margin call, which entails the risk of liquidation of the hedging position. In the second model, the financing is available but a given credit spread is to be paid for this, so hedging can become costly.

Value creation and appropriation through the Hungarian SME’s food chain

Value creation is the result of the continuous innovation activity of the entrepreneur, which is carried out mainly in form of open innovation among the agri-food SMEs. However value creation is not the ultimate goal of the enterprises. They are more interested in increased appropriation of the created value. Although the value creation (innovation) is very well explored and cultivated area of research, there are some voids in the field of agriculture and food industry: the behavioural aspect of open innovation is very rare. The value capturing is even much less studied, therefor our research approach is largely explorative one. Data are drawn from a survey carried out in Hungary among the agri-food SMEs in 2014. We use Structural Equation Modelling as well as ordered probit and semi-non parametric ordered probit models for analysing the data. Our results show that there is positive relationship between the knowledge sharing with chain partners and the innovativeness. We could explore that size of the firm, absorptive capacity and openness to foreign trade ambiguously affects value capturing. However trust in chain partners, reciprocity in knowledge sharing with chain partners and willingness to cooperate with buyers positively influence the appropriation of the created value.

An approach for design and analysis of User-centric Communication Middleware

Today, the development of domain-specific communication applications is both time-consuming and error-prone because the low-level communication services provided by the existing systems and networks are primitive and often heterogeneous. Multimedia communication applications are typically built on top of low-level network abstractions such as TCP/UDP socket, SIP (Session Initiation Protocol) and RTP (Real-time Transport Protocol) APIs. The User-centric Communication Middleware (UCM) is proposed to encapsulate the networking complexity and heterogeneity of basic multimedia and multi-party communication for upper-layer communication applications. And UCM provides a unified user-centric communication service to diverse communication applications ranging from a simple phone call and video conferencing to specialized communication applications like disaster management and telemedicine. It makes it easier to the development of domain-specific communication applications. The UCM abstraction and API is proposed to achieve these goals. The dissertation also tries to integrate the formal method into UCM development process. The formal model is created for UCM using SAM methodology. Some design errors are found during model creation because the formal method forces to give the precise description of UCM. By using the SAM tool, formal UCM model is translated to Promela formula model. In the dissertation, some system properties are defined as temporal logic formulas. These temporal logic formulas are manually translated to promela formulas which are individually integrated with promela formula model of UCM and verified using SPIN tool. Formal analysis used here helps verify the system properties (for example multiparty multimedia protocol) and dig out the bugs of systems.