Abstract

We present a system employing domain specific dictionaries and features to recognize chemical entities. The system utilizes sentence segmentation, tokenization, feature generation, Conditional Random Field (CRF) training and one post-processing step. The dictio-naries were compiled from PubChem, Wikipedia, ChEMBL, DrugBank,word2vec clusters from US patents belonging to A61K class. We report the evaluation results of the run where development set was not included as part of the training set. The best performing model for CEMP taskh as the micro average precision, recall and F-score values of87.26%,79.98%and83.46%, respectively.

Abstract

Sindhi is an Indo-Aryan language spoken by more than 58 million speakers around the world. It is currently a resource poor language which is harmed by the literature being written in multiple scripts. Though the language is widely spoken,primarily, across two countries, the written form is not standardized. In this paper, we seek to develop resources for basic language processing for Sindhi language, in one of its preferred scripts (Devanagari), because a language that seeks to survive in the modern information society requires language technology products. This paper presents our work on building a stochastic Part-of-Speech tagger for Sindhi-Devanagari using conditional random fields with linguistically motivated features. The paper also discusses the steps taken to construct a part-of-speech annotated corpus for Sindhi in Devanagari script. We have also explained in detail the features that were used for training the tagger, which resulted in a part of speech tagger nearing 92% average accuracy.

Abstract

Code-Mixing (CM) is defined as the embedding of linguistic units such as phrases, words, and morphemes of one language into an utterance of another language. CM is a natural phenomenon observed in many multilingual societies. It helps in speeding-up communication and allows wider variety of expression due to which it has become a popular mode of communication in social media forums like Facebook and Twitter. However, current Question Answering (QA) re-search and systems only support expressing a question in a single language which is an unrealistic and hard proposition especially for certain domains like health and technology. In this paper, we take the first step towards the development of a full-fledged QA system in CM language which is building a Question Classification (QC) system. The QC system analyzes the user question and infers the expected Answer Type (AType). The AType helps in locating and verifying the answer as it imposes certain type-specific constraints.We learn a basic Support Vector Machine (SVM) based QC system for English-Hindi CM questions. Due to the inherent complexities involved in processing CM language and also the unavailability of language processing resources such POS taggers, Chunkers, Parsers, we design our current system using only word-level resources such as language iden-tification, transliteration and lexical translation. To reducedata sparsity and leverage resources available in a resource-rich language, in stead of extracting features directly from the original CM words, we translate them commonly into English and then perform featurization. We created an evaluation dataset for this task and our system achieves an ac-curacy of 63% and 45% in coarse-grained and fine-grained categories of the question taxonomy. The idea of translating features into English indeed helps in improving accuracy over the uni-gram baseline

Abstract

This paper presents a rule-based reorder-ing approach for English-Hindi machinetranslation. We have used the concept ofpada,from P ̄an.inian Grammar to framethe reordering rules. Apadais a wordform which is ready to participate in asentence. The rules are generic enoughto apply on any English-Indian languagepair. We tested the rules on English-Hindilanguage pair and obtained better compre-hensibility score as compared to GoogleTranslate on the same test set. In assess-ing the effectiveness of the rules on padaswhich are analogous to minimal phrases inEnglish, we achieved upto 93% accuracyon the test data.

Abstract

Attribute information in a natural language query is one of the key features for converting a natural language query into a Structured Query Language1 (SQL) in Natural Language Interface to Database systems. In this paper, we explore the task of classifying the attributes present in a natural language query into different SQL clauses in a SQL query. In particular, we investigate the effectiveness of various features and Conditional Random Fields for this task. Our system uses a statistical classifier trained on manually prepared data. We report our results on three different domains and also show how our system can be used for generating a complete SQL query.

Abstract

Case-based teaching is a well-known teaching methodology consisting of learning by reading, discussing and analyzing real-life cases and scenarios. We present a Case-Oriented Learning Environment (COSEEd) for teaching Software Engineering concepts to undergraduate and graduate students in a first course of Software Engineering. The novelty of the proposed model lies in being a complete learning environment framework, consisting of pedagogy, broad level learning objectives, assessment, resources and management details, all designed specifically for Software Engineering. Learning and teaching is centered around well-designed SE case studies from authentic software development instances. We describe the COSEEd model, a sample case-study and share out insights as well as lessons learnt while applying the proposed model in practice. We implement and evaluate the proposed model in Software Engineering courses at a University in India focused on the core areas of Information Technology. We use empirical studies on student perception and actual performance to determine the effectiveness of COSEEd towards achieving various learning goals of SE.

Abstract

For the task assignment problem in an expert crowdsourcing platform, we propose that the dynamically arriving workers report their preferences for the tasks as ordinal preferences to the platform. We model then the task assignment problem as a dynamic two sided matching problem. In this paper we study the dynamic two sided matching when the men (the workers) side of the market is arriving dynamically and the women (the requesters) side is available since beginning. We assume strict preferences of the agents. Using a deferred acceptance algorithm as a building block, we first develop f AP ODA, a class of strategy-proof online mechanisms. We design f AP ODA and f T hODA in this class. As no mechanism can achieve stability in our settings, we propose a weaker notion of stability, namely, progressive stability. We introduce an online mechanism f RODA that achieve the progressive stability. For achieving good rank-efficiency, we design an online matching mechanism f BOMA. We study all the four mechanisms empirically for stability and rank-efficiency.

Abstract

Crowdsourcing marketplaces link large populations of workers to an even larger number of tasks. Thus, it is necessary to have mechanisms for matching workers with interesting and suitable tasks. Earlier work has addressed the problem of finding optimal workers for a given set of tasks. However, workers also have preferences and will stay with a platform only if it gives them interesting tasks. We therefore analyze several matching mechanisms that take into account workers’ preferences as well. We propose that the workers pay premiums to get preferred matches and auction-based models where preferences are expressed through variations of the payment for a task. We analyze the properties of two matching different mechanisms: Split Dynamic VCG (SDV) and e-Auction. We compare both the mechanisms with Arrival Priority Serial Dictatorship (APSD) empirically for efficiency.

Abstract

This paper presents a framework for Robust Infrastructure over Shared Computing resource (RISC), which can offer Organizations with Small-scale Computing infrastructures (OSCs) a way to share their unused resources in an ad-hoc manner for suitable monetary incentives. Such a framework provides dual benefits to an OSC: it enables sharing of unused resource during periods of low computing load while allowing execution of any long-term computation on public or anonymized data at a very low cost during periods of high load. The ad-hoc and heterogeneous nature of the shared infrastructure make the resource management problem in RISC non-trivial—a resource manager needs to: (i) maximize profit while determining incentives for resource owners and prices for resource users in an integrated manner; and (ii) emulate large-scale cloud-like robustness and capabilities out of unreliable, small-scale and intermittently available resources at a low cost. This leads to a constrained market situation where offered prices and incentives should lead to a desired level of SLA and reliability for the consumers. Existing approaches of incentive based scheduling for market-like grids assume an open market, based only on demand response; and thus are inapplicable for the constrained market situation in shared resources infrastructure. Specifically, RISC framework has two main components: (i) a firstof-a-kind Dynamic Pricing and Incentivization (DPI) strategy that computes the incentives and the prices while maximizing profit for RISC, using an epoch-by-epoch pricing feedback loop; and (ii) a DPI dependent Reliability, Cost and SLAaware (RCS) scheduler that takes the resource reservation requests as input and assigns replicas of these requests to one or more shared resources for guaranteeing performance SLAs and reliability, while minimizing the cost of resource reservations. Moreover, to handle the communication overhead of computing over geographically distributed resources, the scheduler strives to reduce the network cost of resource allocation. Results from extensive trace-driven experimentation show that our approach can indeed provide appropriate incentives for resource providers, and robust cost-efficient infrastructure solution for resource users.

Abstract

This paper describes the system that was submitted to SemEval2015 Task 10: Sentiment Analysis in Twitter. We participated in Subtask B: Message Polarity Classification. The task is a message level classification of tweets into positive, negative and neutral sentiments. Our model is primarily a supervised one which consists of well designed features fed into an SVM classifier. In previous runs of this task, it was found that lexicons played an important role in determining the sentiment of a tweet. We use existing lexicons to extract lexicon specific features. The lexicon based features are further augmented by tweet specific features. We also improve our system by using acronym and emoticon dictionaries. The proposed system achieves an F1 score of 59.83 and 67.04 on the Test Data and Progress Data respectively. This placed us at the 18th position for the Test Dataset and the 16th position for the Progress Test Dataset.