http://www.bibsonomy.org/burst/user/hotho/toreadBibSonomy BuRST Feed for /user/hotho/toread2008-10-07T18:57:55+02:00http://www.bibsonomy.org/bibtex/257de9154de9e4848eb5989f9ca7fdcbb/hothohotho2008-09-26T16:06:49+02:00toread semantic wordnet similarity distributional <span style="color:#555555;">James Richard <a href="http://www.bibsonomy.org/author/Curran">Curran</a> </span><em>Institute for Communicating and Collaborative Systems School of Informatics University of Edinburgh, </em>(<em>2003</em>)Fri Sep 26 16:06:49 CEST 2008{From Distributional to Semantic Similarity}2003toread semantic wordnet similarity distributional Lexical-semantic resources, including thesauri and WOR DNE T, have been successfully incor- porated into a wide range of applications in Natural Language Processing. However they are very difficult and expensive to create and maintain, and their usefulness has been severely hampered by their limited coverage, bias and inconsistency. Automated and semi-automated methods for developing such resources are therefore crucial for further resource development and improved application performance. Systems that extract thesauri often identify similar words using the distributional hypothesis that similar words appear in similar contexts. This approach involves using corpora to examine the contexts each word appears in and then calculating the similarity between context distri- butions. Different definitions of context can be used, and I begin by examining how different types of extracted context influence similarity. To be of most benefit these systems must be capable of finding synonyms for rare words. Reliable context counts for rare events can only be extracted from vast collections of text. In this dissertation I describe how to extract contexts from a corpus of over 2 billion words. I describe techniques for processing text on this scale and examine the trade-off between context accuracy, information content and quantity of text analysed. Distributional similarity is at best an approximation to semantic similarity. I develop improved approximations motivated by the intuition that some events in the context distribution are more indicative of meaning than others. For instance, the object-of-verb context wear is far more indicative of a clothing noun than get. However, existing distributional techniques do not effectively utilise this information. The new context-weighted similarity metric I propose in this dissertation significantly outperforms every distributional similarity metric described in the literature. Nearest-neighbour similarity algorithms scale poorly with vocabulary and context vector size. To overcome this problem I introduce a new context-weighted approximation algorithm with bounded complexity in context vector size that significantly reduces the system runtime with only a minor performance penalty. I also describe a parallelized version of the system that runs on a Beowulf cluster for the 2 billion word experiments. To evaluate the context-weighted similarity measure I compare ranked similarity lists against gold-standard resources using precision and recall-based measures from Information Retrieval, since the alternative, application-based evaluation, can often be influenced by distributional as well as semantic similarity. I also perform a detailed analysis of the final results using WOR DNE T. Finally, I apply my similarity metric to the task of assigning words to WOR DNE T semantic categories. I demonstrate that this new approach outperforms existing methods and overcomes some of their weaknesses. SpringerLink - Buchkapitelhttp://www.bibsonomy.org/bibtex/227c7357d3337d890fef53168dce9ed33/hothohotho2008-09-11T12:03:43+02:00analysis query search ml dm intention toread <span style="color:#555555;">Ricardo <a href="http://www.bibsonomy.org/author/Baeza-Yates">Baeza-Yates</a> und Liliana <a href="http://www.bibsonomy.org/author/Calderón-Benavides">Calder&#243;n-Benavides</a> und Cristina <a href="http://www.bibsonomy.org/author/González-Caro">Gonz&#225;lez-Caro</a> </span><em>String Processing and Information Retrieval</em>(<em>2006</em>)Thu Sep 11 12:03:43 CEST 2008String Processing and Information Retrieval98--109The Intention Behind Web Queries2006analysis query search ml dm intention toread The identification of the user’s intention or interest through queries that they submit to a search engine can be very useful to offer them more adequate results. In this work we present a framework for the identification of user’s interest in an automaticway, based on the analysis of query logs. This identification is made from two perspectives, the objectives or goals of auser and the categories in which these aims are situated. A manual classification of the queries was made in order to havea reference point and then we applied supervised and unsupervised learning techniques. The results obtained show that fora considerable amount of cases supervised learning is a good option, however through unsupervised learning we found relationshipsbetween users and behaviors that are not easy to detect just taking the query words. Also, through unsupervised learning weestablished that there are categories that we are not able to determine in contrast with other classes that were not consideredbut naturally appear after the clustering process. This allowed us to establish that the combination of supervised and unsupervisedlearning is a good alternative to find user’s goals. From supervised learning we can identify the user interest given certainestablished goals and categories; on the other hand, with unsupervised learning we can validate the goals and categories used,refine them and select the most appropriate to the user’s needs.Tagging and searchinghttp://www.bibsonomy.org/bibtex/27e1dc3f52085093cc33d8fe931253b34/hothohotho2008-09-04T12:37:59+02:00performance engine retrieval toread search ir folksonomy comparision <span style="color:#555555;">P. <a href="http://www.bibsonomy.org/author/Jason Morrison">Jason Morrison</a> </span><em>Inf. Process. Manage.</em><em>44(4):1562--1579</em>(<em>2008</em>)Thu Sep 04 12:37:59 CEST 2008Tarrytown, NY, USAInf. Process. Manage.41562--1579Tagging and searching: Search retrieval effectiveness of folksonomies on the World Wide Web442008performance engine retrieval toread search ir folksonomy comparision Many Web sites have begun allowing users to submit items to a collection and tag them with keywords. The folksonomies built from these tags are an interesting topic that has seen little empirical research. This study compared the search information retrieval (IR) performance of folksonomies from social bookmarking Web sites against search engines and subject directories. Thirty-four participants created 103 queries for various information needs. Results from each IR system were collected and participants judged relevance. Folksonomy search results overlapped with those from the other systems, and documents found by both search engines and folksonomies were significantly more likely to be judged relevant than those returned by any single IR system type. The search engines in the study had the highest precision and recall, but the folksonomies fared surprisingly well. Del.icio.us was statistically indistinguishable from the directories in many cases. Overall the directories were more precise than the folksonomies but they had similar recall scores. Better query handling may enhance folksonomy IR performance further. The folksonomies studied were promising, and may be able to improve Web search performance.[cond-mat/0508028] A generative model for feedback networkshttp://www.bibsonomy.org/bibtex/2dd7cd33e8a95a0128fe05adc46483ac7/hothohotho2008-08-28T10:19:29+02:00network model toread <span style="color:#555555;">Douglas R. <a href="http://www.bibsonomy.org/author/White">White</a> und Natasa <a href="http://www.bibsonomy.org/author/Kejzar">Kejzar</a> und Constantino <a href="http://www.bibsonomy.org/author/Tsallis">Tsallis</a> und Doyne <a href="http://www.bibsonomy.org/author/Farmer">Farmer</a> und Scott <a href="http://www.bibsonomy.org/author/White">White</a> </span>(<em>2005</em>)Thu Aug 28 10:19:29 CEST 2008A generative model for feedback networks2005network model toread We investigate a simple generative model for network formation. The model is designed to describe the growth of networks of kinship, trading, corporate alliances, or autocatalytic chemical reactions, where feedback is an essential element of network growth. The underlying graphs in these situations grow via a competition between cycle formation and node addition. After choosing a given node, a search is made for another node at a suitable distance. If such a node is found, a link is added connecting this to the original node, and increasing the number of cycles in the graph; if such a node cannot be found, a new node is added, which is linked to the original node. We simulate this algorithm and find that we cannot reject the hypothesis that the empirical degree distribution is a q-exponential function, which has been used to model long-range processes in nonequilibrium statistical mechanics.http://www.bibsonomy.org/bibtex/2534bb3931ebf4a537d7d4e3c85788632/hothohotho2008-08-17T21:16:05+02:00toread <span style="color:#555555;">Jun <a href="http://www.bibsonomy.org/author/Zhu">Zhu</a> und Zaiqing <a href="http://www.bibsonomy.org/author/Nie">Nie</a> und Bo <a href="http://www.bibsonomy.org/author/Zhang">Zhang</a> und Ji-Rong <a href="http://www.bibsonomy.org/author/Wen">Wen</a> </span><em>v9, </em>(<em>1583</em>)Sun Aug 17 21:16:05 CEST 2008Dynamic Hierarchical Markov Random Fields for Integrated Web Data Extraction v91583toread Lillian Lee's Home Pagehttp://www.bibsonomy.org/bibtex/2236d4f703fda3dd9457863f28eda56cb/hothohotho2008-08-13T10:34:04+02:00toread sentiment mining opinion analysis <span style="color:#555555;">Bo <a href="http://www.bibsonomy.org/author/Pang">Pang</a> und Lillian <a href="http://www.bibsonomy.org/author/Lee">Lee</a> </span><em>Foundations and Trends&#174; in Information Retrieval</em><em>2(1-2):1-135</em>(<em>2008</em>)Wed Aug 13 10:34:04 CEST 2008Foundations and Trends® in Information Retrieval1-21-135Opinion mining and sentiment analysis22008toread sentiment mining opinion analysis [0710.2889] An efficient reduction of ranking to classificationhttp://www.bibsonomy.org/bibtex/2d8bd1b99e3c245d17b577514727ebff2/hothohotho2008-07-15T18:12:46+02:00toread learning ranking <span style="color:#555555;">Nir <a href="http://www.bibsonomy.org/author/Ailon">Ailon</a> und Mehryar <a href="http://www.bibsonomy.org/author/Mohri">Mohri</a> </span>(<em>2007</em>)Tue Jul 15 18:12:46 CEST 2008An efficient reduction of ranking to classification2007toread learning ranking This paper describes an efficient reduction of the learning problem of ranking to binary classification. The reduction guarantees an average pairwise misranking regret of at most that of the binary classifier regret, improving a recent result of Balcan et al which only guarantees a factor of 2. Moreover, our reduction applies to a broader class of ranking loss functions, admits a simpler proof, and the expected running time complexity of our algorithm in terms of number of calls to a classifier or preference function is improved from $\Omega(n^2)$ to $O(n \log n)$. In addition, when the top $k$ ranked elements only are required ($k \ll n$), as in many applications in information extraction or search engines, the time complexity of our algorithm can be further reduced to $O(k \log k + n)$. Our reduction and algorithm are thus practical for realistic applications where the number of points to rank exceeds several thousands. Much of our results also extend beyond the bipartite case previously studied.http://www.bibsonomy.org/bibtex/2645abd6b3191a2a6e844d7542651ed1c/hothohotho2008-07-12T12:19:07+02:00clusterig folksonomy detection toread community <span style="color:#555555;">Akshay <a href="http://www.bibsonomy.org/author/Java">Java</a> und Anupam <a href="http://www.bibsonomy.org/author/Joshi">Joshi</a> und Tim <a href="http://www.bibsonomy.org/author/Finin">Finin</a> </span><em>WebKDD 2008 Workshop on Web Mining and Web Usage Analysis, </em><em>August2008. </em><em>To Appear . </em>Sat Jul 12 12:19:07 CEST 2008WebKDD 2008 Workshop on Web Mining and Web Usage AnalysisAugustTo Appear{Detecting Commmunities via Simultaneous Clustering of Graphs and Folksonomies}2008clusterig folksonomy detection toread community Welcome to IEEE Xplore 2.0: CoolRank: A Social Solution for Ranking Bookmarked Web Resourceshttp://www.bibsonomy.org/bibtex/24671fb1c606e3d7f559bb25d9b20e47d/hothohotho2008-07-02T16:57:41+02:00folksonomy 2.0 * toread ranking web folkrank <span style="color:#555555;">H.S. <a href="http://www.bibsonomy.org/author/Al-Khalifa">Al-Khalifa</a> </span><em>Innovations in Information Technology, 2007. Innovations '07. 4th International Conference on, </em><em>Seite208-212. </em>(<em>2007</em>)Wed Jul 02 16:57:41 CEST 2008Innovations in Information Technology, 2007. Innovations '07. 4th International Conference on208-212CoolRank: A Social Solution for Ranking Bookmarked Web Resources2007folksonomy 2.0 * toread ranking web folkrank Users tag resources for a variety of reasons and using a variety of conventions. The tags that they provide are stored in social bookmarking services, so these services can provide a rich gateway to a wide and interesting quantity of web resources. The cognitive effort that has gone into making these tags has presumably added value to the description of the resource. In this work we utilize the quantitative value of these tags for ranking bookmarked web resources in social bookmarking services. Our proposed solution is called CoolRank, a simple and intuitive model to rank bookmarked web resources in a social bookmarking service, such as del.icio.us. CoolRank makes use of both quantitative information, based on the number of people who have bookmarked a web resource, and subjective information, based on the words people have used in their tags.http://www.bibsonomy.org/bibtex/252943a6298169f5a552bffbbee352937/hothohotho2008-06-12T08:19:04+02:00toread iswc social search bookmarking tagging <span style="color:#555555;">Michael <a href="http://www.bibsonomy.org/author/Noll">Noll</a> und Christoph <a href="http://www.bibsonomy.org/author/Meinel">Meinel</a> </span><em>Proceedings of the 6th International Semantic Web Conference and 2nd Asian Semantic Web Conference (ISWC/ASWC2007), Busan, South Korea, </em><em>Volume4825vonLNCS, </em><em>Seite365--378. </em><em>Berlin, Heidelberg, </em><em>Springer Verlag, </em><em>November2007. </em>Thu Jun 12 08:19:04 CEST 2008Berlin, HeidelbergProceedings of the 6th International Semantic Web Conference and 2nd Asian Semantic Web Conference (ISWC/ASWC2007), Busan, South Koreahttp://data.semanticweb.org/conference/iswc-aswc/2007/proceedingsNovember365--378LNCSWeb search personalization via social bookmarking and tagging48252007toread iswc social search bookmarking tagging In this paper, we present a new approach to web search personalization based on user collaboration and sharing of information about web documents. The proposed personalization technique separates data collection and user profiling from the information system whose contents and indexed documents are being searched for, i.e. the search engines, and uses social bookmarking and tagging to re-rank web search results. It is independent of the search engine being used, so users are free to choose the one they prefer, even if their favorite search engine does not natively support personalization. We show how to design and implement such a system in practice and investigate its feasibility and usefulness with large sets of real-word data and a user study.Institut AIFB - Publikation: Ontologies on Demand? - A Description of the State-of-the-Art, Applications, Challenges and Trends for Ontology Learning from Texthttp://www.bibsonomy.org/bibtex/2fe4c2950b5be221b493e29e4339240e8/hothohotho2008-05-20T17:43:51+02:00toread survey learning ontology <span style="color:#555555;">Philipp <a href="http://www.bibsonomy.org/author/Cimiano">Cimiano</a> und Johanna <a href="http://www.bibsonomy.org/author/Völker">V&#246;lker</a> und Rudi <a href="http://www.bibsonomy.org/author/Studer">Studer</a> </span><em>Information, Wissenschaft und Praxis</em><em>57(6-7):315-320</em><em>OCT2006. </em><em>see the special issue for more contributions related to the Semantic Web . </em>Tue May 20 17:43:51 CEST 2008Information, Wissenschaft und PraxisOCTsee the special issue for more contributions related to the Semantic Web6-7315-320Ontologies on Demand? - A Description of the State-of-the-Art, Applications, Challenges and Trends for Ontology Learning from Text572006toread survey learning ontology Semantic feature production norms for a large set ...[Behav Res Methods. 2005] - PubMed Resulthttp://www.bibsonomy.org/bibtex/2936af12b025e37b0a6aac6bc103f58a3/hothohotho2008-05-08T12:17:01+02:00dataset semantic toread ol ontology relation grounding <span style="color:#555555;">K <a href="http://www.bibsonomy.org/author/McRae">McRae</a> und G S <a href="http://www.bibsonomy.org/author/Cree">Cree</a> und M S <a href="http://www.bibsonomy.org/author/Seidenberg">Seidenberg</a> und C <a href="http://www.bibsonomy.org/author/McNorgan">McNorgan</a> </span><em>Behav Res Methods</em><em>37(4):547-559</em><em>Nov2005. </em>Thu May 08 12:17:01 CEST 2008Behav Res MethodsNov4547-559Semantic feature production norms for a large set of living and nonliving things372005dataset semantic toread ol ontology relation grounding Semantic features have provided insight into numerous behavioral phenomena concerning concepts, categorization, and semantic memory in adults, children, and neuropsychological populations. Numerous theories and models in these areas are based on representations and computations involving semantic features. Consequently, empirically derived semantic feature production norms have played, and continue to play, a highly useful role in these domains. This article describes a set of feature norms collected from approximately 725 participants for 541 living (dog) and nonliving (chair) basic-level concepts, the largest such set of norms developed to date. This article describes the norms and numerous statistics associated with them. Our aim is to make these norms available to facilitate other research, while obviating the need to repeat the labor-intensive methods involved in collecting and analyzing such norms. The full set of norms may be downloaded from www.psychonomic.org/archive.BibSonomy::edit bibtexhttp://www.bibsonomy.org/bibtex/276d1018ba398695e454d20de302de6e6/hothohotho2008-05-06T09:43:15+02:00clustering SVM toread code <span style="color:#555555;">Sei-Hyung <a href="http://www.bibsonomy.org/author/Lee">Lee</a> und Karen M. <a href="http://www.bibsonomy.org/author/Daniels">Daniels</a> </span><em>Proceedings of 6th SIAM Conference on Data Mining, </em><em>Seite484&#8211;488. </em><em>May2006. </em>Tue May 06 09:43:15 CEST 2008Proceedings of 6th SIAM Conference on Data MiningMay484–488Cone Cluster Labeling for Support Vector Clustering2006clustering SVM toread code March 2008http://www.bibsonomy.org/bibtex/27a080f640fa62fc81e73b9fab1e7447c/hothohotho2008-04-26T13:11:10+02:00flickr toread dynamics social network digg <span style="color:#555555;">Kristina <a href="http://www.bibsonomy.org/author/Lerman">Lerman</a> </span><em>arXiv</em><em>Jan2007. </em>Sat Apr 26 13:11:10 CEST 2008arXivJanSocial Information Processing in Social News Aggregation2007flickr toread dynamics social network digg The rise of the social media sites, such as blogs, wikis, Digg and Flickr among others, underscores the transformation of the Web to a participatory medium in which users are collaboratively creating, evaluating and distributing information. The innovations introduced by social media has lead to a new paradigm for interacting with information, what we call 'social information processing'. In this paper, we study how social news aggregator Digg exploits social information processing to solve the problems of document recommendation and rating. First, we show, by tracking stories over time, that social networks play an important role in document recommendation. The second contribution of this paper consists of two mathematical models. The first model describes how collaborative rating and promotion of stories emerges from the independent decisions made by many users. The second model describes how a user's influence, the number of promoted stories and the user's social network, changes in time. We find qualitative agreement between predictions of the model and user data gathered from Digg.A framework for community identification in dynamic social networkshttp://www.bibsonomy.org/bibtex/227a4fb58300979d4dbe94e75422418bd/hothohotho2008-04-26T12:29:32+02:00toread clustering graph community detection <span style="color:#555555;">Chayant <a href="http://www.bibsonomy.org/author/Tantipathananandh">Tantipathananandh</a> und Tanya <a href="http://www.bibsonomy.org/author/Berger-Wolf">Berger-Wolf</a> und David <a href="http://www.bibsonomy.org/author/Kempe">Kempe</a> </span><em>KDD '07: Proceedings of the 13th ACM SIGKDD international conference on Knowledge discovery and data mining, </em><em>Seite717--726. </em><em>New York, NY, USA, </em><em>ACM, </em>(<em>2007</em>)Sat Apr 26 12:29:32 CEST 2008New York, NY, USAKDD '07: Proceedings of the 13th ACM SIGKDD international conference on Knowledge discovery and data mining717--726A framework for community identification in dynamic social networks2007toread clustering graph community detection http://www.bibsonomy.org/bibtex/25e5cc221d7da719909f3bf8c507b0afc/hothohotho2008-04-24T09:02:43+02:00model mining graph toread <span style="color:#555555;">D. <a href="http://www.bibsonomy.org/author/Chakrabarti">Chakrabarti</a> und Y. <a href="http://www.bibsonomy.org/author/Zhan">Zhan</a> und C. <a href="http://www.bibsonomy.org/author/Faloutsos">Faloutsos</a> </span><em>SIAM International Conference on Data Mining, </em>(<em>2004</em>)Thu Apr 24 09:02:43 CEST 2008SIAM International Conference on Data MiningR-MAT: A Recursive Model for Graph Mining2004model mining graph toread DBLP Record 'conf/iis/OsinskiSW04'http://www.bibsonomy.org/bibtex/240aba631c1ac8819bd64b0ee74bfdd1b/hothohotho2008-03-21T15:24:47+01:00svd clustering toread lsi <span style="color:#555555;">Stanislaw <a href="http://www.bibsonomy.org/author/Osinski">Osinski</a> und Jerzy <a href="http://www.bibsonomy.org/author/Stefanowski">Stefanowski</a> und Dawid <a href="http://www.bibsonomy.org/author/Weiss">Weiss</a> </span><em>Intelligent Information Systems, </em><em>Seite359-368. </em>(<em>2004</em>)Fri Mar 21 15:24:47 CET 2008Intelligent Information SystemsConfIis2004359-368Lingo: Search Results Clustering Algorithm Based on Singular Value Decomposition2004svd clustering toread lsi Novelty and collective attention -- Wu and Huberman 104 (45): 17599 -- Proceedings of the National Academy of Scienceshttp://www.bibsonomy.org/bibtex/2ff0a7c4758b8bfdf5cf117f652884728/hothohotho2008-03-18T08:14:03+01:00toread folksonomy attention collective novelty <span style="color:#555555;">F. <a href="http://www.bibsonomy.org/author/Wu">Wu</a> und B. A. <a href="http://www.bibsonomy.org/author/Huberman">Huberman</a> </span><em>Proc. Natl. Acad. Sci. USA</em><em>104(45):17599-17601</em>(<em>2007</em>)Tue Mar 18 08:14:03 CET 2008Proc. Natl. Acad. Sci. USA4517599-17601Novelty and collective attention1042007toread folksonomy attention collective novelty The subject of collective attention is central to an information age where millions of people are inundated with daily messages. It is thus of interest to understand how attention to novel items propagates and eventually fades among large populations. We have analyzed the dynamics of collective attention among 1 million users of an interactive web site, digg.com, devoted to thousands of novel news stories. The observations can be described by a dynamical model characterized by a single novelty factor. Our measurements indicate that novelty within groups decays with a stretched-exponential law, suggesting the existence of a natural time scale over which attention fades. Approximating the Community Structure of the Long Tailhttp://www.bibsonomy.org/bibtex/2386f36679c111f30e37ced272d5b355c/hothohotho2008-02-19T10:24:54+01:00toread clustering svd detection community <span style="color:#555555;">Akshay <a href="http://www.bibsonomy.org/author/Java">Java</a> und Anupam <a href="http://www.bibsonomy.org/author/Joshi">Joshi</a> und Tim <a href="http://www.bibsonomy.org/author/FininBook">FininBook</a> </span><em>Proceedings of the Second International Conference on Weblogs and Social Media(ICWSM 2008), </em><em>AAAI Press, </em>(<em>2008</em>)Tue Feb 19 10:24:54 CET 2008Proceedings of the Second International Conference on Weblogs and Social Media(ICWSM 2008)Approximating the Community Structure of the Long Tail2008toread clustering svd detection community In many social media applications, a small fraction of the members are highly linked while most are sparsely connected to the network. Such a skewed distribution is sometimes referred to as the"long tail". Popular applications like meme trackers and content aggregators mine for information from only the popular blogs located at the head of this curve. On the other hand, the long tail contains large volumes of interesting information and niches. The question we address in this work is how best to approximate the community membership of entities in the long tail using only a small percentage of the entire graph structure. Our technique utilizes basic linear algebra manipulations and spectral methods. It has the advantage of quickly and efficiently finding a reasonable approximation of the community structure of the overall network. Such a method has significant applications in blog analysis engines as well as social media monitoring tools in general. ScienceDirect - Technological Forecasting and Social Change : Literature-related discovery (LRD): Potential treatments for cataractshttp://www.bibsonomy.org/bibtex/2b9359f79985da9b9677340ffda849e74/hothohotho2007-12-30T11:00:21+01:00clustering discovery toread text semantic retrieval mining <span style="color:#555555;">Ronald N. <a href="http://www.bibsonomy.org/author/Kostoff">Kostoff</a> </span><em>Technological Forecasting and Social Change</em>(<em>2007</em>)Sun Dec 30 11:00:21 CET 2007Technological Forecasting and Social Change--Literature-related discovery (LRD): Potential treatments for cataractsIn Press, Corrected Proof2007clustering discovery toread text semantic retrieval mining Literature-related discovery (LRD) is the linking of two or more literature concepts that have heretofore not been linked (i.e., disjoint), in order to produce novel, interesting, plausible, and intelligible knowledge (i.e., potential discovery). The open discovery systems (ODS) component of LRD starts with a problem to be solved, and generates solutions to that problem through potential discovery. We have been using ODS LRD to identify potential treatments or preventative actions for challenging medical problems, among myriad other applications. This paper describes the second medical problem we addressed (cataract) using ODS LRD; the first problem addressed was Raynaud's Phenomenon (RP), and was described in the third paper of this Special Issue. Cataract was selected because it is ubiquitous globally, appears intractable to all forms of treatment other than surgical removal of cataracts, and is a major cause of blindness in many developing countries. The ODS LRD study had three objectives: a) identify non-drug non-surgical treatments that would 1) help prevent cataracts, or 2) reduce the progression rate of cataracts, or 3) stop the progression of cataracts, or 4) maybe even reverse the progression of cataracts; b) demonstrate that we could solve an ODS LRD problem with no prior knowledge of any results or prior work (unlike the case with the RP problem); c) determine whether large time savings in the discovery process were possible relative to the time required for conducting the RP study. To that end, we used the MeSH taxonomy of MEDLINE to restrict potential discoveries to selected semantic classes, as a substitute for the manually-intensive process used in the RP study to restrict potential discoveries to selected semantic classes. We also used additional semantic filtering to identify potential discovery within the selected semantic classes. All these goals were achieved. As will be shown, we generated large amounts of potential discovery in more than an order of magnitude less time than required for the RP study. We identified many non-drug non-surgical treatments that may be able to reduce or even stop the progression rate of cataracts. Time, and much testing, will determine whether this is possible. Finally, the methodology has been developed to the point where ODS LRD problems can be solved with no results or knowledge of any prior work.