Production of Referring Expressions for an Unknown Audience: A Computational Model of Communal Common Ground

This article presents a computational model of the production of referring expressions under uncertainty over the hearer's knowledge. Although situations where the hearer's knowledge is uncertain have seldom been addressed in the computational literature, they are common in ordinary communication, for example when a writer addresses an unknown audience, or when a speaker addresses a stranger. We propose a computational model composed of three complimentary heuristics based on, respectively, an estimation of the recipient's knowledge, an estimation of the extent to which a property is unexpected, and the question of what is the optimum number of properties in a given situation. The model was tested in an experiment with human readers, in which it was compared against the Incremental Algorithm and human-produced descriptions. The results suggest that the new model outperforms the Incremental Algorithm in terms of the proportion of correctly identified entities and in terms of the perceived quality of the generated descriptions.

Crowdsourcing Without a Crowd

We present an incremental Bayesian model that resolves key issues of crowd size and data quality for consensus labeling. We evaluate our method using data collected from a real-world citizen science program, B ee W atch , which invites members of the public in the United Kingdom to classify (label) photographs of bumblebees as one of 22 possible species. The biological recording domain poses two key and hitherto unaddressed challenges for consensus models of crowdsourcing: (1) the large number of potential species makes classification difficult, and (2) this is compounded by limited crowd availability, stemming from both the inherent difficulty of the task and the lack of relevant skills among the general public. We demonstrate that consensus labels can be reliably found in such circumstances with very small crowd sizes of around three to five users (i.e., through group sourcing). Our incremental Bayesian model, which minimizes crowd size by re-evaluating the quality of the consensus label following each species identification solicited from the crowd, is competitive with a Bayesian approach that uses a larger but fixed crowd size and outperforms majority voting. These results have important ecological applicability: biological recording programs such as B ee W atch can sustain themselves when resources such as taxonomic experts to confirm identifications by photo submitters are scarce (as is typically the case), and feedback can be provided to submitters in a timely fashion. More generally, our model provides benefits to any crowdsourced consensus labeling task where there is a cost (financial or otherwise) associated with soliciting a label.

The role of automated feedback in training and retaining biological recorders for citizen science

The rapid rise of citizen science, with lay people forming often extensive biodiversity sensor networks, is seen as a solution to the mismatch between data demand and supply while simultaneously engaging citizens with environmental topics. However, citizen science recording schemes require careful consideration of how to motivate, train, and retain volunteers. We evaluated a novel computing science framework that allowed for the automated generation of feedback to citizen scientists using natural language generation (NLG) technology. We worked with a photo-based citizen science program in which users also volunteer species identification aided by an online key. Feedback is provided after photo (and identification) submission and is aimed to improve volunteer species identification skills and to enhance volunteer experience and retention. To assess the utility of NLG feedback, we conducted two experiments with novices to assess short-term (single session) and longer-term (5 sessions in 2 months) learning, respectively. Participants identified a specimen in a series of photos. One group received only the correct answer after each identification, and the other group received the correct answer and NLG feedback explaining reasons for misidentification and highlighting key features that facilitate correct identification. We then developed an identification training tool with NLG feedback as part of the citizen science program BeeWatch and analyzed learning by users. Finally, we implemented NLG feedback in the live program and evaluated this by randomly allocating all BeeWatch users to treatment groups that received different types of feedback upon identification submission. After 6 months separate surveys were sent out to assess whether views on the citizen science program and its feedback differed among the groups. Identification accuracy and retention of novices were higher for those who received automated feedback than for those who received only confirmation of the correct identification without explanation. The value of NLG feedback in the live program, captured through questionnaires and evaluation of the online photo-based training tool, likewise showed that the automated generation of informative feedback fostered learning and volunteer engagement and thus paves the way for productive and long-lived citizen science projects.

Automated data analysis to rapidly derive and communicate ecological insights from satellite-tag data: a case study of reintroduced red kites

Analysis of satellite-telemetry data mostly occurs long after it has been collected, due to the time and effort needed to collate and interpret such material. Delayed reporting reduces the usefulness of such data for nature conservation where timely information about animal movements is required. To counter this problem, we present a novel approach which combines automated analysis of satellite-telemetry data with rapid communication of insights derived from such data. A relatively simple algorithm (based on radial and angular velocity calculated from fixes) allowed instantaneous detection of excursions away from settlement areas and automated calculation of home ranges on the remaining data. Automating the detection of both excursions and home-range calculations enabled us to disseminate ecological insights from satellite-tag data instantaneously through a dedicated web portal. The automated analysis, interpretation, and communication of satellite-tag and other ecological data offer clear benefits to nature conservation research and practice.

Mapping species distributions: a comparison of skilled naturalist and lay citizen science recording

To assess the ability of traditional biological recording schemes and lay citizen science approaches to gather data on species distributions and changes therein, we examined bumblebee records from the UK's national repository (National Biodiversity Network) and from BeeWatch. The two recording approaches revealed similar relative abundances of bumblebee species but different geographical distributions. For the widespread common carder (Bombus pascuorum), traditional recording scheme data were patchy, both spatially and temporally, reflecting active record centre rather than species distribution. Lay citizen science records displayed more extensive geographic coverage, reflecting human population density, thus offering better opportunities to account for recording effort. For the rapidly spreading tree bumblebee (Bombus hypnorum), both recording approaches revealed similar distributions due to a dedicated mapping project which overcame the patchy nature of naturalist records. We recommend, where possible, complementing skilled naturalist recording with lay citizen science programmes to obtain a nation-wide capability, and stress the need for timely uploading of data to the national repository.

Basic Gene Grammars and DNA-ChartParser for language processing of Escherichia coli promoter DNA sequences

MOTIVATION

The field of 'DNA linguistics' has emerged from pioneering work in computational linguistics and molecular biology. Most formal grammars in this field are expressed using Definite Clause Grammars but these have computational limitations which must be overcome. The present study provides a new DNA parsing system, comprising a logic grammar formalism called Basic Gene Grammars and a bidirectional chart parser DNA-ChartParser.

RESULTS

The use of Basic Gene Grammars is demonstrated in representing many formulations of the knowledge of Escherichia coli promoters, including knowledge acquired from human experts, consensus sequences, statistics (weight matrices), symbolic learning, and neural network learning. The DNA-ChartParser provides bidirectional parsing facilities for BGGs in handling overlapping categories, gap categories, approximate pattern matching, and constraints. Basic Gene Grammars and the DNA-ChartParser allowed different sources of knowledge for recognizing E.coli promoters to be combined to achieve better accuracy as assessed by parsing these DNA sequences in real-world data sets.