Báo cáo khoa học: "Splitting Complex Temporal Questions for Question Answering systems" doc

... primary?”. TE: 1988 Complex Temporal Questions: Type 3: Multiple events temporal questions with temporal expression. Questions that contain two ormore events, related by a temporal signal. This ... Decompo-sition Unit for the treatment of complex questions. For the evaluation a corpus of questions containingas many simple as complex questions is required.Due to the fact that question corpora ... work presents a Question Answering systemcapable of answering complex temporal questions. This approach tries to imitate human behavior whenresponding this type of questions. For example, ahuman...

Báo cáo khoa học: "A Modal Temporal Logic for Reasoning about Change" ppt

... then p is a formula. - If p and q are formulas, then (-p), (p v q) are formulas. - If m is a unary temporal operator and p is a formula, then (m p) is a formula. - If m is a binary temporal ... often needs to provide more than just the literal answer to a question. For example, [Kaplan 82I presents methods for correcting a questionerls misconceptions (as reflected in a query) about ... in the former it is. This paper is concerned with developing a lo~ic for reasoning about change in data bases, and assessing the impact of that capability on the behavior of question answering...

Tài liệu Báo cáo khoa học: "Phrase-Based Translation Model for Question Retrieval in Community Question Answer Archives" ppt

... al., 2008). Therefore, it is a meaningfultask to retrieve the questions that are semanticallyequivalent or relevant to the queried questions. For example in Table 1, given question Q1, Q2can ... between the queried questions andthe historical questions in the archives (Jeon et al.,2005; Xue et al., 2008). For example in Table 1, Q1and Q2are two semantically similar questions, butthey ... aims to ﬁnd historical questions thatare semantically equivalent or relevant to thequeried questions. In this paper, we proposea novel phrase-based translation model for question retrieval. Compared...

Tài liệu Báo cáo khoa học: "Generating Impact-Based Summaries for Scientiﬁc Literature" docx

... our methodconsistently outperforms all the baselines. Amongthe baselines, MEAD -Doc is consistently better thanboth LEAD and MEAD -Doc+ Cite. While MEAD- Doc s outperforming LEAD is not surprising, ... surprising that MEAD -Doc also outperformsMEAD -Doc+ Cite as the latter uses both the cita-tion context and the original document. One possi-ble explanation may be that MEAD is not designed for impact ... using thedocument collection.Table 2: Impact-based summary vs. regular summary for the paper “A study of smoothing methods for languagemodels applied to ad hoc information retrieval”.nal document...

Tài liệu Báo cáo khoa học: "Mining Wikipedia Revision Histories for Improving Sentence Compression" docx

... many NLP tasks, it is much more severe for sentence compression, leading Turner and Charniak(2005) to question the applicability of the channelmodel for this task altogether.Our contribution ... of them for such compres-sions/expansions. We make the simplifying assump-tion that all such edits also retain the core mean-ing of the sentence, and are therefore valid trainingdata for our ... for glareprotection] is effective and will help if youroffice has the fluorescent-light overkill [that’s typical in offices].(4) Prices range from $5,000 [for a microvax2000] to $179,000 [for...

Tài liệu Báo cáo khoa học: "Exploiting N-best Hypotheses for SMT Self-Enhancement" doc

... Association for Computational LinguisticsExploiting N-best Hypotheses for SMT Self-Enhancement Boxing Chen, Min Zhang, Aiti Aw and Haizhou Li Department of Human Language Technology Institute for ... Chinese-to-English task show performance improvements for all the strategies. Moreover, the combination of the three strategies achieves further improvements and outperforms the baseline by 0.67 ... than those that did not. We call the former “good phrase-pairs” and the later “bad phrase-pairs” for the given source input. Hypothetically, we can reinforce the translation model by appending...

Tài liệu Báo cáo khoa học: "Context Management with Topics for Spoken Dialogue Systems" pptx

... average precision for all the pruned topic types is 74.64%, varying from 95.63% for ROOM to 37.63% for MIx. If MIx is left out, the average precision is 79.27%. The poor precision for MIX is due ... the fact that the information content of the utterances is ignored. Consistency of the information exchanged in (task- oriented) conversations is one of the main sources for dialogue coherence, ... heuristics for different shifts of attention, we seek a formalisation of the information structure of utterances in terms of the new information that is exchanged in the course of the dialogue....

Tài liệu Báo cáo khoa học: "ADP based Search Algorithm for Statistical Machine Translation" docx

... additional parameter into the recursion formula for DP. In the following, we will explain this method in detail. 2.3 Recursion Formula for DP In the DP formalism, the search process is described ... little meaningful information or the information is different from the input. Examples for each category are given in Table 3. Table 4 shows the statistics of the translation performance. When ... reduces the time complexity by a factor J. 2. Optimizing the target string length as formu- lated in Eq. (10) requires the dynamic program- ming procedure to start all over again for each I....

Tài liệu Báo cáo khoa học: "AN INTEGRATED HEURISTIC SCHEME FOR PARTIAL PARSE EVALUATION" docx

... fea- tures are designed to be general and, for the most part, grammar and domain independent. For each parse, the heuristic computes a penalty score for each of the fea- tures. The penalties ... AN INTEGRATED HEURISTIC SCHEME FOR PARTIAL PARSE EVALUATION Alon Lavie School of Computer Science Carnegie Mellon University 5000 Forbes Ave., Pittsburgh, PA 15213 email : ... tree The penalty scheme for skipped words is designed to prefer parses that correspond to fewer skipped words. It assigns a penalty in the range of (0.95 - 1.05) for each word of the original...

Báo cáo khoa học: "Sentence Dependency Tagging in Online Question Answering Forums" pdf

... post. Only identifying question and answering sentences from the thread is not enough for automatic question answering. For this example,in order to get the complete question, we would needto ... the question and answer sentences from forumthreads. However, there is much richer informationin forum conversations, and simply knowing a sen-tence is a question or answer is not enough. For example, ... answers. Question answering forums, suchas technical support forums, are places where usersﬁnd answers through conversations. Because oftheir nature as online communities, question answer-ing forums...

Xem thêm