Results 1 -
8 of
8
Discriminative models of integrating document evidence and document-candidate associations for expert search
- In Proceeding of the 33rd international ACM SIGIR conference on Research and development in information retrieval, SIGIR ’10
, 2010
"... Generative models such as statistical language modeling have been widely studied in the task of expert search to model the relationship between experts and their expertise indi-cated in supporting documents. On the other hand, dis-criminative models have received little attention in expert search re ..."
Abstract
-
Cited by 14 (1 self)
- Add to MetaCart
(Show Context)
Generative models such as statistical language modeling have been widely studied in the task of expert search to model the relationship between experts and their expertise indi-cated in supporting documents. On the other hand, dis-criminative models have received little attention in expert search research, although they have been shown to outper-form generative models in many other information retrieval and machine learning applications. In this paper, we propose a principled relevance-based discriminative learning frame-work for expert search and derive specific discriminative models from the framework. Compared with the state-of-the-art language models for expert search, the proposed re-search can naturally integrate various document evidence and document-candidate associations into a single model without extra modeling assumptions or effort. An extensive set of experiments have been conducted on two TREC En-terprise track corpora (i.e., W3C and CERC) to demonstrate the effectiveness and robustness of the proposed framework.
Combining Candidate and Document Models for Expert Search
"... Abstract: We describe our participation in the TREC 2008 Enterprise track and detail our language modeling-based approaches. For document search, our focus was on query expansion using profiles of top ranked experts and on document priors. We found that these techniques result in small, but noticeab ..."
Abstract
-
Cited by 6 (1 self)
- Add to MetaCart
(Show Context)
Abstract: We describe our participation in the TREC 2008 Enterprise track and detail our language modeling-based approaches. For document search, our focus was on query expansion using profiles of top ranked experts and on document priors. We found that these techniques result in small, but noticeable improvements over our baseline method. For expert search, we combine candidate- and document-based models, and also bring in web evidence. We found that the combined models significantly and consistently outperformed our very competitive baseline models. 1
Adapting language modeling methods for expert search to rank Wikipedia entities
- Lecture Notes in Computer Science
, 2009
"... Abstract. In this paper, we propose two methods to adapt language modeling methods for expert search to the INEX entity ranking task. In our experiments, we notice that language modeling methods for expert search, if directly applied to the INEX entity ranking task, cannot effectively distinguish en ..."
Abstract
-
Cited by 4 (0 self)
- Add to MetaCart
(Show Context)
Abstract. In this paper, we propose two methods to adapt language modeling methods for expert search to the INEX entity ranking task. In our experiments, we notice that language modeling methods for expert search, if directly applied to the INEX entity ranking task, cannot effectively distinguish entity types. Thus, our proposed methods aim at resolving this problem. First, we propose a method to take into account the INEX category query field. Second, we use an interpolation of two language models to rank entities, which can solely work on the text query. Our experiments indicate that both methods can effectively adapt language modeling methods for expert search to the INEX entity ranking task.
CSIR at TREC 2008 expert search task: Modeling expert evidence in expert search
- In Proceedings of the 2008 Text REtrieval Conference (TREC 2008
, 2008
"... Abstract. In this paper, we described our method for the expert search task in TREC 2008. First, we proposed an adaption to the language modeling method for expert search, which considers the probability of query generation separately using each kind of expert evidence (full name, abbreviated name, ..."
Abstract
-
Cited by 2 (0 self)
- Add to MetaCart
(Show Context)
Abstract. In this paper, we described our method for the expert search task in TREC 2008. First, we proposed an adaption to the language modeling method for expert search, which considers the probability of query generation separately using each kind of expert evidence (full name, abbreviated name, and email address). Current expert search models can be easily integrated into our method. Our experiments indicated that our method can make use of the ambiguous evidence in expert search (abbreviated name), which often casued a drop in effects in other methods. Besides, we also used a probabilistic measure to detect phrase in query, but it did not make better effectiveness.
Blog, Enterprise
"... Abstract: We describe the participation of the University of Amsterdam’s ILPS group in the blog, enterprise and relevance feedback track at TREC 2008. Our main preliminary conclusions are that estimating mixture weights for external expansion in blog post retrieval is non-trivial and we need more an ..."
Abstract
- Add to MetaCart
Abstract: We describe the participation of the University of Amsterdam’s ILPS group in the blog, enterprise and relevance feedback track at TREC 2008. Our main preliminary conclusions are that estimating mixture weights for external expansion in blog post retrieval is non-trivial and we need more analysis to find out why it works better for blog distillation than for blog post retrieval. For the relevance feedback track we observe two things: (i) in terms of statMAP, a larger number of judged non-relevant documents improves retrieval effectiveness and (ii) on the TREC Terabyte topics, we can effectively replace the estimates on the judged non-relevant documents with estimations on the document collection. Finally, since the enterprise track did not have any results yet, we only described our participation and do not draw any conclusions. 1
Expert Search on Web Using Association Distribution
"... Abstract — Diverse environments have studied expert search viz. educational communities, enterprises etc. The system refers to a universal expert search problem: Most important is expert search on the internet, that considering ordinary WebPages and people names. But it is having primarily two chall ..."
Abstract
- Add to MetaCart
Abstract — Diverse environments have studied expert search viz. educational communities, enterprises etc. The system refers to a universal expert search problem: Most important is expert search on the internet, that considering ordinary WebPages and people names. But it is having primarily two challenging issues: Unreliable quality of WebPages and WebPages containing full of unwanted information; usually indistinct and confusing expertise evidence are spread in web pages. To address the task of finding experts on the web, numerous solutions have been proposed. Relevance is the main concern in usual organizational expert search, so it is essential to take advantage of the huge amount of co-occurrence information to evaluate relevance and reputation of an individual name for a query theme. This paper mainly proposes a multithreaded ranking algorithm which considers people names and ordinary web pages. We are complementing both document and proximity-based approaches to expert finding by importing global evidence of expertise. Our proposed system also deals with the problem of extraction and disambiguation of person name. An NLP technique to adjust association scores among people and words is also applied by proposed system.
Concomitant Supported Dissemination for Expert Search on Web
"... Expert search has been studied in various contexts. We inspect a commonexpert search problem. Piercing experts on the web, where thousands of names and millions of webpages are considered. It has two challenging issues: 1) web pages could be of full of noises and low quality2) The expertise evidence ..."
Abstract
- Add to MetaCart
(Show Context)
Expert search has been studied in various contexts. We inspect a commonexpert search problem. Piercing experts on the web, where thousands of names and millions of webpages are considered. It has two challenging issues: 1) web pages could be of full of noises and low quality2) The expertise evidences scattered in web pages are usually arises confusion. We propose to leverage the large amount of concomitant information to assess relevance and reputation of a person name for a query topic. The concomitant structure is modeled using a hyper graph, on which a heat disseminates based ranking algorithm is proposed. Query keywords are regarded as sources of heat and a name of person which has strong link with the query (i.e., frequently
