| { | |
| "paper_id": "W10-0402", | |
| "header": { | |
| "generated_with": "S2ORC 1.0.0", | |
| "date_generated": "2023-01-19T05:04:49.370156Z" | |
| }, | |
| "title": "Scientific Authoring Support: A Tool to Navigate in Typed Citation Graphs", | |
| "authors": [ | |
| { | |
| "first": "Ulrich", | |
| "middle": [], | |
| "last": "Sch\u00e4fer", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "Language Technology Lab German Research Center for Artificial Intelligence (DFKI)", | |
| "institution": "", | |
| "location": { | |
| "postCode": "D-66123", | |
| "settlement": "Saarbr\u00fccken", | |
| "country": "Germany" | |
| } | |
| }, | |
| "email": "[email protected]" | |
| }, | |
| { | |
| "first": "Uwe", | |
| "middle": [], | |
| "last": "Kasterka", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "", | |
| "institution": "Saarland University", | |
| "location": { | |
| "addrLine": "Campus D", | |
| "postCode": "66123", | |
| "settlement": "Saarbr\u00fccken", | |
| "country": "Germany" | |
| } | |
| }, | |
| "email": "[email protected]" | |
| } | |
| ], | |
| "year": "", | |
| "venue": null, | |
| "identifiers": {}, | |
| "abstract": "Scientific authors urgently need help in managing the fast increasing number of publications. We describe and demonstrate a tool that supports authors in browsing graphically through electronically available publications, thus allowing them to quickly adapt to new domains and publish faster. Navigation is assisted by means of typed citation graphs, i.e. we use methods and resources from computational linguistics to compute the kind of citation that is made from one paper to another (refutation, use, confirmation etc.). To verify the computed citation type, the user can inspect the highlighted citation sentence in the original PDF document. While our classification methods used to generate a realistic test data set are relatively simple and could be combined with other proposed approaches, we put a strong focus on usability and quick navigation in the potentially huge graphs. In the outlook, we argue that our tool could be made part of a community approach to overcome the sparseness and correctness dilemma in citation classification.", | |
| "pdf_parse": { | |
| "paper_id": "W10-0402", | |
| "_pdf_hash": "", | |
| "abstract": [ | |
| { | |
| "text": "Scientific authors urgently need help in managing the fast increasing number of publications. We describe and demonstrate a tool that supports authors in browsing graphically through electronically available publications, thus allowing them to quickly adapt to new domains and publish faster. Navigation is assisted by means of typed citation graphs, i.e. we use methods and resources from computational linguistics to compute the kind of citation that is made from one paper to another (refutation, use, confirmation etc.). To verify the computed citation type, the user can inspect the highlighted citation sentence in the original PDF document. While our classification methods used to generate a realistic test data set are relatively simple and could be combined with other proposed approaches, we put a strong focus on usability and quick navigation in the potentially huge graphs. In the outlook, we argue that our tool could be made part of a community approach to overcome the sparseness and correctness dilemma in citation classification.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Abstract", | |
| "sec_num": null | |
| } | |
| ], | |
| "body_text": [ | |
| { | |
| "text": "According to different studies, the number of scientific works is doubled every 5-10 years. Important issues to be addressed by the scientific community are finding relevant information and avoiding redundancy and duplication of work. The organization and preservation of scientific knowledge in scientific publications, vulgo text documents, thwarts these efforts. From a viewpoint of a computer scientist, scientific papers are just 'unstructured information'.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Introduction and Motivation", | |
| "sec_num": "1" | |
| }, | |
| { | |
| "text": "One specific, but very important aspect of the content of scientific papers is their relation to previous work and, once published, their impact to subsequent or derived research. While it is still hard if not impossible to capture and formalize the semantic content of a scientific publication automatically, at least citation properties and derived scientific impact can be and usually are measured automatically on the basis of simple citation graphs. In other words, these graphs can be used to describe I/O behavior of publications in a very simple way.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Introduction and Motivation", | |
| "sec_num": "1" | |
| }, | |
| { | |
| "text": "However, just counting citations is a very coarse approach and does not tell much about the reasons for citing one's work in a specific situation. Moreover, once such measure is formalized and standardized e.g. for science evaluation, it can be exploited to tune up statistics. Since the first proposal of the Science Citation Index (Garfield, 1955) , it has also provoked criticism.", | |
| "cite_spans": [ | |
| { | |
| "start": 333, | |
| "end": 349, | |
| "text": "(Garfield, 1955)", | |
| "ref_id": "BIBREF7" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Introduction and Motivation", | |
| "sec_num": "1" | |
| }, | |
| { | |
| "text": "In the bibliometrics and computational linguistics literature, many proposals are available on how citations could be further classified by careful analysis of citation sentences and context (Garfield, 1965; Garzone, 1996; Mercer and Di Marco, 2004; Teufel et al., 2006; Bornmann and Daniel, 2008) .", | |
| "cite_spans": [ | |
| { | |
| "start": 191, | |
| "end": 207, | |
| "text": "(Garfield, 1965;", | |
| "ref_id": "BIBREF8" | |
| }, | |
| { | |
| "start": 208, | |
| "end": 222, | |
| "text": "Garzone, 1996;", | |
| "ref_id": "BIBREF9" | |
| }, | |
| { | |
| "start": 223, | |
| "end": 249, | |
| "text": "Mercer and Di Marco, 2004;", | |
| "ref_id": "BIBREF11" | |
| }, | |
| { | |
| "start": 250, | |
| "end": 270, | |
| "text": "Teufel et al., 2006;", | |
| "ref_id": "BIBREF17" | |
| }, | |
| { | |
| "start": 271, | |
| "end": 297, | |
| "text": "Bornmann and Daniel, 2008)", | |
| "ref_id": "BIBREF3" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Introduction and Motivation", | |
| "sec_num": "1" | |
| }, | |
| { | |
| "text": "The number of different classes proposed varies from 3 to 35. Different authors try to identify dimensions and mutually exclusive classes, but the more classes a schema contains, the more difficult becomes the automatic classification.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Introduction and Motivation", | |
| "sec_num": "1" | |
| }, | |
| { | |
| "text": "The focus of our paper is to combine automatic classification approaches with a tool that supports scientists in graphically navigating through typed citation graphs (TCG). Such TCGs can be generated by augmenting a simple citation graph with information synonymously called citation function (Teufel et al., 2006) , citation relation (Mercer and Di Marco, 2004) or citation sentiment, forming the labels of the graph's edges. In the following, we use the more neutral and general term citation type.", | |
| "cite_spans": [ | |
| { | |
| "start": 293, | |
| "end": 314, | |
| "text": "(Teufel et al., 2006)", | |
| "ref_id": "BIBREF17" | |
| }, | |
| { | |
| "start": 335, | |
| "end": 362, | |
| "text": "(Mercer and Di Marco, 2004)", | |
| "ref_id": "BIBREF11" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Introduction and Motivation", | |
| "sec_num": "1" | |
| }, | |
| { | |
| "text": "The idea is to help scientists, especially those not so familiar with an area, understanding the relations between publications and quickly get an overview of the field. Moreover, the goal is to embed this tool in a larger framework for scientists that also supports semantic search assisted by domain ontologies and further tools for authoring support (Sch\u00e4fer et al., 2008) .", | |
| "cite_spans": [ | |
| { | |
| "start": 353, | |
| "end": 375, | |
| "text": "(Sch\u00e4fer et al., 2008)", | |
| "ref_id": "BIBREF16" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Introduction and Motivation", | |
| "sec_num": "1" | |
| }, | |
| { | |
| "text": "Our paper is structured as follows. In Section 2, we describe how we automatically compute the typed citation graph from the raw text content of a scientific paper corpus to generate realistic data for testing the visualization and navigation tool. Section 3 contains an evaluation of the quality of the extracted unlabeled graphs and of the citation classification step. We then describe in Section 4 the ideas of efficient and at the same time well-arranged visualization and navigation in the typed citation graph. We compare with related work in Section 5. Finally, we conclude and give an outlook to future work in Section 6.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Introduction and Motivation", | |
| "sec_num": "1" | |
| }, | |
| { | |
| "text": "Our corpus is based on 6300 electronically-available papers, a subset (published 2002-2008) of the ACL Anthology (Bird et al., 2008) , a comprehensive collection of scientific conference and workshop papers in the area of computational linguistics and language technology. The overall workflow of the employed tools and data is shown in Fig. 1 .", | |
| "cite_spans": [ | |
| { | |
| "start": 70, | |
| "end": 91, | |
| "text": "(published 2002-2008)", | |
| "ref_id": null | |
| }, | |
| { | |
| "start": 113, | |
| "end": 132, | |
| "text": "(Bird et al., 2008)", | |
| "ref_id": "BIBREF2" | |
| } | |
| ], | |
| "ref_spans": [ | |
| { | |
| "start": 337, | |
| "end": 343, | |
| "text": "Fig. 1", | |
| "ref_id": "FIGREF0" | |
| } | |
| ], | |
| "eq_spans": [], | |
| "section": "Data Preparation and Automatic Citation Type Classification", | |
| "sec_num": "2" | |
| }, | |
| { | |
| "text": "We ran the open source tool ParsCit (Councill et al., 2008) to extract references lists and corresponding citation sentences from raw paper texts. To build the citation graph, we used the Levenshtein distance (Levenshtein, 1966) to find and match titles and authors of identical papers yet tolerating spelling and PDF extraction errors.", | |
| "cite_spans": [ | |
| { | |
| "start": 36, | |
| "end": 59, | |
| "text": "(Councill et al., 2008)", | |
| "ref_id": "BIBREF5" | |
| }, | |
| { | |
| "start": 209, | |
| "end": 228, | |
| "text": "(Levenshtein, 1966)", | |
| "ref_id": "BIBREF10" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Data Preparation and Automatic Citation Type Classification", | |
| "sec_num": "2" | |
| }, | |
| { | |
| "text": "To increase robustness, publication years were not considered as they would hinder matches for delayed journal publications. Generation of the graph edges, i.e. matching of papers and reference strings, is performed by means of the ACL ID, a unique identifier for each paper, available for the PDF (source nodes of references) and BibTeX files (targets of references).", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Data Preparation and Automatic Citation Type Classification", | |
| "sec_num": "2" | |
| }, | |
| { | |
| "text": "We evaluated the generated graph against the one that was corrected manually by the ACL Anthology Network (AAN) group (Radev et al., 2009) and found that 10821 citation links were shared between both and can be considered correct 1 .", | |
| "cite_spans": [ | |
| { | |
| "start": 118, | |
| "end": 138, | |
| "text": "(Radev et al., 2009)", | |
| "ref_id": "BIBREF15" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Data Preparation and Automatic Citation Type Classification", | |
| "sec_num": "2" | |
| }, | |
| { | |
| "text": "3883 additional ones were in the AAN but not recognized by us, the other way round, 1021 discovered by us were not in the AAN. In addition, the publication bases were not identical. The anthology network data ends in February 2007 but covers years before 2002, while our data covers 2002-2008 inclusively. Given the fact that our graph is computed fully automatically, the result can be considered very good.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Data Preparation and Automatic Citation Type Classification", | |
| "sec_num": "2" | |
| }, | |
| { | |
| "text": "In the next step, we augmented the citation graph by types for each edge. In contrast to other approaches, we currently only consider the citation sentence itself to determine the citation type, neither a wider context, its position nor the abstract, title or content of the cited paper. A reference (from the references section at the end of a paper) may be associated with several citation sentences mentioning the paper referenced at the end.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Data Preparation and Automatic Citation Type Classification", | |
| "sec_num": "2" | |
| }, | |
| { | |
| "text": "In only considering the citation sentence itself, we may lose some citation type information, as it may be (also) contained in follow-up sentences referring to the citation using a pronoun (\"they\", \"their approach\" etc.). Considering follow-up or even preceding sentences is planned to be addressed in future work.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Data Preparation and Automatic Citation Type Classification", | |
| "sec_num": "2" | |
| }, | |
| { | |
| "text": "After consulting the rich literature on citation classification (Bornmann and Daniel, 2008; Garzone, 1996; Teufel et al., 2006) , we derived a simplified classification schema consisting of the following five classes.", | |
| "cite_spans": [ | |
| { | |
| "start": 64, | |
| "end": 91, | |
| "text": "(Bornmann and Daniel, 2008;", | |
| "ref_id": "BIBREF3" | |
| }, | |
| { | |
| "start": 92, | |
| "end": 106, | |
| "text": "Garzone, 1996;", | |
| "ref_id": "BIBREF9" | |
| }, | |
| { | |
| "start": 107, | |
| "end": 127, | |
| "text": "Teufel et al., 2006)", | |
| "ref_id": "BIBREF17" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Data Preparation and Automatic Citation Type Classification", | |
| "sec_num": "2" | |
| }, | |
| { | |
| "text": "\u2022 Agree: The citing paper agrees with the cited paper", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Data Preparation and Automatic Citation Type Classification", | |
| "sec_num": "2" | |
| }, | |
| { | |
| "text": "\u2022 PRecycle: The citing paper uses an algorithm, data, method or tool from the cited paper", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Data Preparation and Automatic Citation Type Classification", | |
| "sec_num": "2" | |
| }, | |
| { | |
| "text": "\u2022 Negative:", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Data Preparation and Automatic Citation Type Classification", | |
| "sec_num": "2" | |
| }, | |
| { | |
| "text": "The paper is cited negatively/contrastively", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Data Preparation and Automatic Citation Type Classification", | |
| "sec_num": "2" | |
| }, | |
| { | |
| "text": "\u2022 Neutral: The paper is cited neutrally", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Data Preparation and Automatic Citation Type Classification", | |
| "sec_num": "2" | |
| }, | |
| { | |
| "text": "\u2022 Undef: impossible determine the sentiment of the citation (fallback)", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Data Preparation and Automatic Citation Type Classification", | |
| "sec_num": "2" | |
| }, | |
| { | |
| "text": "Then, we used a blend of methods to collect verbal and non-verbal patterns (cue words) and associated each with a class from the aforementioned schema.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Data Preparation and Automatic Citation Type Classification", | |
| "sec_num": "2" | |
| }, | |
| { | |
| "text": "\u2022 A list from (Garzone, 1996) devised for biomedical texts; it is largely applicable to the computational linguistics domain as well.", | |
| "cite_spans": [ | |
| { | |
| "start": 14, | |
| "end": 29, | |
| "text": "(Garzone, 1996)", | |
| "ref_id": "BIBREF9" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Data Preparation and Automatic Citation Type Classification", | |
| "sec_num": "2" | |
| }, | |
| { | |
| "text": "\u2022 Simple negation of positive cue words to obtain negative patterns.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Data Preparation and Automatic Citation Type Classification", | |
| "sec_num": "2" | |
| }, | |
| { | |
| "text": "\u2022 A list of automatically extracted synonyms and antonyms (the latter for increasing number of patterns for negative citations) from WordNet (Miller et al., 1993 ).", | |
| "cite_spans": [ | |
| { | |
| "start": 141, | |
| "end": 161, | |
| "text": "(Miller et al., 1993", | |
| "ref_id": "BIBREF12" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Data Preparation and Automatic Citation Type Classification", | |
| "sec_num": "2" | |
| }, | |
| { | |
| "text": "\u2022 Automatically computed most frequent cooccurrences from all extracted citation sentences of the corpus using an open source cooccurrence tool (Banerjee and Pedersen, 2003) .", | |
| "cite_spans": [ | |
| { | |
| "start": 144, | |
| "end": 173, | |
| "text": "(Banerjee and Pedersen, 2003)", | |
| "ref_id": "BIBREF0" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Data Preparation and Automatic Citation Type Classification", | |
| "sec_num": "2" | |
| }, | |
| { | |
| "text": "\u2022 Inspection: browse and filter cue words manually, remove redundancies.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Data Preparation and Automatic Citation Type Classification", | |
| "sec_num": "2" | |
| }, | |
| { | |
| "text": "These pattern where then used for the classification algorithm and applied to the extracted citation sentences. In case of multiple citations with different classes, a voting mechanism was applied were the 'stronger' classes (Agree, Negative, PRecycle) won in standoff cases. For the total of 91419 citations we obtained the results shown in The numbers reflect a careful classification approach where uncertain citations are classified as Undef. In case of multiple matches, the first (leftmost) was taken to achieve a unique result.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Results: Distribution and Evaluation", | |
| "sec_num": "3" | |
| }, | |
| { | |
| "text": "The results also confirm obervations made in other works: (1) citation classification is a hard task, (2) there are only a few strongly negative citations which coincides with observations made by (Teufel et al., 2006) , (Pendlebury, 2009) and others, (3) the majority of citations is neutral or of unknown type.", | |
| "cite_spans": [ | |
| { | |
| "start": 197, | |
| "end": 218, | |
| "text": "(Teufel et al., 2006)", | |
| "ref_id": "BIBREF17" | |
| }, | |
| { | |
| "start": 221, | |
| "end": 239, | |
| "text": "(Pendlebury, 2009)", | |
| "ref_id": "BIBREF14" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Results: Distribution and Evaluation", | |
| "sec_num": "3" | |
| }, | |
| { | |
| "text": "An evaluation on a test set of 100 citations spread across all the types of papers with a manual check of the accuracy of the computed labels showed an overall accuracy of 30% mainly caused by the fact that 90% of undefined hits were in fact neutral (i.e., labeling all undefs neutral would increase accuracy). Negative citations are sparse and unreliable (33%), neutral ones are about 60% accurate, PRecycle: 33%, Agree: 25%.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Results: Distribution and Evaluation", | |
| "sec_num": "3" | |
| }, | |
| { | |
| "text": "To sum up, our automatic classification approach based on only local citation information could surely be improved by applying methods described in the literature, but it helped us to quickly (without annotation effort) generate a plausible data set for the main task, visualization and navigation in the typed citation graphs. ", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Results: Distribution and Evaluation", | |
| "sec_num": "3" | |
| }, | |
| { | |
| "text": "The overall idea of the citation visualization and navigation tool is simple and intuitive. Each paper is represented by a node, all citations between papers are drawn as edges between nodes where the color of the edge indicates the computed (overall) citation type, e.g. green for agree, red for negative, blue for recycle and black for neutral or undefined.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Visualization Algorithm and Navigation User Interface", | |
| "sec_num": "4" | |
| }, | |
| { | |
| "text": "To cope with flexible layouts and scalability of the graph, we decided to use the open source tool Java Universal Network/Graph Framework (JUNG, http://jung.sourceforge.net). Its main advantages over similar tools are that it supports user interaction (clicking on nodes and edges, tool tips) and userimplemented graph layout algorithms. A screenshot of the final user interface is presented in Figure 2 .", | |
| "cite_spans": [], | |
| "ref_spans": [ | |
| { | |
| "start": 395, | |
| "end": 403, | |
| "text": "Figure 2", | |
| "ref_id": "FIGREF1" | |
| } | |
| ], | |
| "eq_spans": [], | |
| "section": "Visualization Algorithm and Navigation User Interface", | |
| "sec_num": "4" | |
| }, | |
| { | |
| "text": "The decision for and development of the visualization and navigation tool was mainly driven by the fact that citation graphs quickly grow and become unmanagable by humans when extended to the transitive closures of citing or cited papers of a given publication. The sheer number of crossing edges would make the display unreadable. The main design goal therefore was reducing the number of nodes and edges where possible and (by default) have only one paper in focus (Fig. 3) , with all cited papers on the left side (Fig. 4) , and all citing papers on the right (Fig. 5) .", | |
| "cite_spans": [], | |
| "ref_spans": [ | |
| { | |
| "start": 467, | |
| "end": 475, | |
| "text": "(Fig. 3)", | |
| "ref_id": "FIGREF2" | |
| }, | |
| { | |
| "start": 517, | |
| "end": 525, | |
| "text": "(Fig. 4)", | |
| "ref_id": "FIGREF3" | |
| }, | |
| { | |
| "start": 563, | |
| "end": 571, | |
| "text": "(Fig. 5)", | |
| "ref_id": "FIGREF4" | |
| } | |
| ], | |
| "eq_spans": [], | |
| "section": "Visualization Algorithm and Navigation User Interface", | |
| "sec_num": "4" | |
| }, | |
| { | |
| "text": "This also reflects a time line order where the origin (oldest papers) is on the left. In the graphical user interface, the citation depth (default 1) is adjustable by a slider to higher numbers. The graph display is updated upon change of the configured depth. At level 1, papers citing the citing papers (analogously for cited papers), are not fully drawn as nodes, but only adumbrated by short ingoing or outgoing edges (arrows). However, the color of these short edges still signifies the citation type and may attract interest which can easily be satisfied by clicking on the edge's remaining node (cf. screenshot in Figure 2 ). When the mouse is moved over a node, a tooltip text display pops up displaying full author list and paper title. To avoid crossing edges caused by citations at the same level (citing or cited papers also cite each other), we devised a fan-out layout generation (Fig. 6) . It increases the width of the displayed graph, but leads to better readability. Fan-out layout can also be switched off in the user interface. In addition, the graph layout algorithm orders papers chronologically in the vertical direction. Here, we have implemented another technique that helps to avoid crossing edges. As shown in Fig. 7 , we sort papers vertically by also taking into account the position of its predecessor, the cited paper. It often leads to less crossing edges. By double-clicking on any node representing a paper (cited or citing), this node can be made the new center and the graph is re-arranged accordingly.", | |
| "cite_spans": [], | |
| "ref_spans": [ | |
| { | |
| "start": 621, | |
| "end": 629, | |
| "text": "Figure 2", | |
| "ref_id": "FIGREF1" | |
| }, | |
| { | |
| "start": 894, | |
| "end": 902, | |
| "text": "(Fig. 6)", | |
| "ref_id": "FIGREF5" | |
| }, | |
| { | |
| "start": 1237, | |
| "end": 1243, | |
| "text": "Fig. 7", | |
| "ref_id": "FIGREF6" | |
| } | |
| ], | |
| "eq_spans": [], | |
| "section": "Visualization Algorithm and Navigation User Interface", | |
| "sec_num": "4" | |
| }, | |
| { | |
| "text": "Zooming in and out is possible via mouse wheel or shortcut buttons ('overview', 'center').", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Visualization Algorithm and Navigation User Interface", | |
| "sec_num": "4" | |
| }, | |
| { | |
| "text": "Using the right mouse button context menu on a node, it is possible to open a details page for the selected paper with bibliographic metadata and all citations and types. All references in the document with their citation sentences identified are displayed in a structured list.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Visualization Algorithm and Navigation User Interface", | |
| "sec_num": "4" | |
| }, | |
| { | |
| "text": "The citation context around a citation sentence is shown as well, while the citation sentence itself is colored according to the citation type color and clickable. If clicked, the original PDF document opens with the citation sentence highlighted ( Fig. 8 ; currently only possible in Acrobat Reader).", | |
| "cite_spans": [], | |
| "ref_spans": [ | |
| { | |
| "start": 249, | |
| "end": 255, | |
| "text": "Fig. 8", | |
| "ref_id": null | |
| } | |
| ], | |
| "eq_spans": [], | |
| "section": "Visualization Algorithm and Navigation User Interface", | |
| "sec_num": "4" | |
| }, | |
| { | |
| "text": "By clicking on an edge instead of a node, only the citations between the two papers at both ends are displayed, in the same way as described above for all citations of a document.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Visualization Algorithm and Navigation User Interface", | |
| "sec_num": "4" | |
| }, | |
| { | |
| "text": "Our paper touches and combines results of three disciplines, (1) bibliometrics, (2) computational linguistics, and (3) information visualization. We briefly discuss related and mostly recent literature, being aware of the fact that this list is necessarily incomplete. (Garfield, 1965) is probably the first to discuss an automatic computation of citation types. He is also the founder of citation indexing and the Institute of Scientific Information (ISI). His first publication on science citation indexing appeared in 1955 (Garfield, 1955) and he remained the probably most influential scientist in this field for decades. (Bornmann and Daniel, 2008 ) is a comprehensive recent metastudy on citing behavior.", | |
| "cite_spans": [ | |
| { | |
| "start": 269, | |
| "end": 285, | |
| "text": "(Garfield, 1965)", | |
| "ref_id": "BIBREF8" | |
| }, | |
| { | |
| "start": 526, | |
| "end": 542, | |
| "text": "(Garfield, 1955)", | |
| "ref_id": "BIBREF7" | |
| }, | |
| { | |
| "start": 626, | |
| "end": 652, | |
| "text": "(Bornmann and Daniel, 2008", | |
| "ref_id": "BIBREF3" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Related Work", | |
| "sec_num": "5" | |
| }, | |
| { | |
| "text": "Investigating citation classification has a long tradition in bibliometrics and information science and in the last 20 years also attracted computational linguistics researchers trying to automate the task based on rhetorics of science, statistical methods and sentence parsing.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Related Work", | |
| "sec_num": "5" | |
| }, | |
| { | |
| "text": "There is much more work than we can cite here on citation function computation worth combination with our approach (Bornmann and Daniel, 2008; Garzone, 1996; Teufel et al., 2006) -using our tool one can easily browse to further publications! There is little work on innovative layout techniques for displaying and navigating citation graphs. We found three independent approaches to citation graph visualization: CiteViz (Elmqvist and Tsigas, 2004) , CircleView (Bergstr\u00f6m and Jr., 2006), and (Nguyen et al., 2007) . They share a disadvantageous property in that they try to visualize too much information at the same time. In our opinion, this contradicts the need to navigate and keep control over displayable parts of large paper collections.", | |
| "cite_spans": [ | |
| { | |
| "start": 115, | |
| "end": 142, | |
| "text": "(Bornmann and Daniel, 2008;", | |
| "ref_id": "BIBREF3" | |
| }, | |
| { | |
| "start": 143, | |
| "end": 157, | |
| "text": "Garzone, 1996;", | |
| "ref_id": "BIBREF9" | |
| }, | |
| { | |
| "start": 158, | |
| "end": 178, | |
| "text": "Teufel et al., 2006)", | |
| "ref_id": "BIBREF17" | |
| }, | |
| { | |
| "start": 421, | |
| "end": 448, | |
| "text": "(Elmqvist and Tsigas, 2004)", | |
| "ref_id": "BIBREF6" | |
| }, | |
| { | |
| "start": 462, | |
| "end": 476, | |
| "text": "(Bergstr\u00f6m and", | |
| "ref_id": "BIBREF1" | |
| }, | |
| { | |
| "start": 477, | |
| "end": 492, | |
| "text": "Jr., 2006), and", | |
| "ref_id": "BIBREF1" | |
| }, | |
| { | |
| "start": 493, | |
| "end": 514, | |
| "text": "(Nguyen et al., 2007)", | |
| "ref_id": "BIBREF13" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Related Work", | |
| "sec_num": "5" | |
| }, | |
| { | |
| "text": "Moreover, these approaches do not provide information on citation types derived from text as our system does. Further ideas on visualizing sciencerelated information such as author co-citation networks are also discussed and summarized in (Chen, 2006) .", | |
| "cite_spans": [ | |
| { | |
| "start": 239, | |
| "end": 251, | |
| "text": "(Chen, 2006)", | |
| "ref_id": "BIBREF4" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Related Work", | |
| "sec_num": "5" | |
| }, | |
| { | |
| "text": "We have presented an innovative tool to support scientific authors in browsing graphically through large collections of publications by means of typed citation graphs. To quickly generate a realistic data set, we devised a classification approach avoiding manual annotation and intervention.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Summary and Outlook", | |
| "sec_num": "6" | |
| }, | |
| { | |
| "text": "Our classification results cannot compete with approaches such as (Teufel et al., 2006 ) based on considerable manual annotation for machine learning. However, we think that our application could be combined with this or other approaches described for classifying citations between scientific papers.", | |
| "cite_spans": [ | |
| { | |
| "start": 66, | |
| "end": 86, | |
| "text": "(Teufel et al., 2006", | |
| "ref_id": "BIBREF17" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Summary and Outlook", | |
| "sec_num": "6" | |
| }, | |
| { | |
| "text": "We envisage to integrate the navigation tool in a larger framework supporting scientific authoring (Sch\u00e4fer et al., 2008) . When publishing a service of this kind on the Web, one would be faced with ethical issues such as the problem that authors could feel offended by wrongly classified citations.", | |
| "cite_spans": [ | |
| { | |
| "start": 99, | |
| "end": 121, | |
| "text": "(Sch\u00e4fer et al., 2008)", | |
| "ref_id": "BIBREF16" | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Summary and Outlook", | |
| "sec_num": "6" | |
| }, | |
| { | |
| "text": "The reason is that citation type classification is potentially even more subjective than a bare citation index-which itself is already highly controversal, as discussed in the introduction. Moreover, there is not always a single, unique citation type, but often vagueness and room for interpretation.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Summary and Outlook", | |
| "sec_num": "6" | |
| }, | |
| { | |
| "text": "Therefore, we suggest to augment such a service by a Web 2.0 application that would allow registered users to confirm, alter and annotate precom- Figure 8 : Citation sentence viewer; citation sentence in context on the left, highlighted in PDF on the right when selected on the left puted citation classifications. In this community application, all citation links in the automatically generated graph could be represented by dashed arrows initially, and users could turn them solid by confirming or correcting the citation type and also adding a comment text.", | |
| "cite_spans": [], | |
| "ref_spans": [ | |
| { | |
| "start": 146, | |
| "end": 154, | |
| "text": "Figure 8", | |
| "ref_id": null | |
| } | |
| ], | |
| "eq_spans": [], | |
| "section": "Summary and Outlook", | |
| "sec_num": "6" | |
| }, | |
| { | |
| "text": "Line thickness could be increased (up to an appropriate maximum) each time another user confirms a classified citation type. The results could then also be employed for active learning and help to improve the automatic classification procedure.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Summary and Outlook", | |
| "sec_num": "6" | |
| }, | |
| { | |
| "text": "We only consider intra-network links here, not those pointing to books or other publications outside the corpus.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "", | |
| "sec_num": null | |
| } | |
| ], | |
| "back_matter": [ | |
| { | |
| "text": "First of all, we are indebted to the anonymous reviewers for their useful, encouraging and detailed comments. Many thanks also to Donia Scott for her feedback on an earlier version of the tool and helpful comments on terminology. We would like to thank Madeline Maher and Boris Fersing for generating and evaluating the citation type data on a subcorpus of the ACL Anthology.The work described in this paper has been carried out in the context of the project TAKE (Technolo-gies for Advanced Knowledge Extraction), funded under contract 01IW08003 by the German Federal Ministry of Education and Research.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Acknowledgments", | |
| "sec_num": null | |
| } | |
| ], | |
| "bib_entries": { | |
| "BIBREF0": { | |
| "ref_id": "b0", | |
| "title": "The design, implementation, and use of the ngram statistics package", | |
| "authors": [ | |
| { | |
| "first": "Satanjeev", | |
| "middle": [], | |
| "last": "Banerjee", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Ted", | |
| "middle": [], | |
| "last": "Pedersen", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2003, | |
| "venue": "Proceedings of the Fourth International Conference on Intelligent Text Processing and Computational Linguistics", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "370--381", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Satanjeev Banerjee and Ted Pedersen. 2003. The de- sign, implementation, and use of the ngram statistics package. In Proceedings of the Fourth International Conference on Intelligent Text Processing and Com- putational Linguistics, pages 370-381, Mexico City.", | |
| "links": null | |
| }, | |
| "BIBREF1": { | |
| "ref_id": "b1", | |
| "title": "Cir-cleView: Scalable visualization and navigation of citation networks", | |
| "authors": [ | |
| { | |
| "first": "Peter", | |
| "middle": [], | |
| "last": "Bergstr\u00f6m", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "E. James", | |
| "middle": [], | |
| "last": "Whitehead", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2006, | |
| "venue": "Proceedings of the 2006 Symposium on Interactive Visual Information Collections and Activity IVICA", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Peter Bergstr\u00f6m and E. James Whitehead Jr. 2006. Cir- cleView: Scalable visualization and navigation of cita- tion networks. In Proceedings of the 2006 Symposium on Interactive Visual Information Collections and Ac- tivity IVICA, College Station, Texas.", | |
| "links": null | |
| }, | |
| "BIBREF2": { | |
| "ref_id": "b2", | |
| "title": "The ACL anthology reference corpus: A reference dataset for bibliographic research", | |
| "authors": [ | |
| { | |
| "first": "Steven", | |
| "middle": [], | |
| "last": "Bird", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Robert", | |
| "middle": [], | |
| "last": "Dale", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Bonnie", | |
| "middle": [], | |
| "last": "Dorr", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Bryan", | |
| "middle": [], | |
| "last": "Gibson", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Mark", | |
| "middle": [], | |
| "last": "Joseph", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Min-Yen", | |
| "middle": [], | |
| "last": "Kan", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Dongwon", | |
| "middle": [], | |
| "last": "Lee", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Brett", | |
| "middle": [], | |
| "last": "Powley", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Dragomir", | |
| "middle": [], | |
| "last": "Radev", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Yee", | |
| "middle": [ | |
| "Fan" | |
| ], | |
| "last": "Tan", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2008, | |
| "venue": "Proceedings of the Language Resources and Evaluation Conference (LREC-2008)", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Steven Bird, Robert Dale, Bonnie Dorr, Bryan Gibson, Mark Joseph, Min-Yen Kan, Dongwon Lee, Brett Powley, Dragomir Radev, and Yee Fan Tan. 2008. The ACL anthology reference corpus: A reference dataset for bibliographic research. In Proceedings of the Lan- guage Resources and Evaluation Conference (LREC- 2008), Marrakesh, Morocco.", | |
| "links": null | |
| }, | |
| "BIBREF3": { | |
| "ref_id": "b3", | |
| "title": "What do citation counts measure? A review of studies on citing behavior", | |
| "authors": [ | |
| { | |
| "first": "Lutz", | |
| "middle": [], | |
| "last": "Bornmann", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Hans-Dieter", | |
| "middle": [], | |
| "last": "Daniel", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2008, | |
| "venue": "Journal of Documentation", | |
| "volume": "64", | |
| "issue": "1", | |
| "pages": "45--80", | |
| "other_ids": { | |
| "DOI": [ | |
| "10.1108/00220410810844150" | |
| ] | |
| }, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Lutz Bornmann and Hans-Dieter Daniel. 2008. What do citation counts measure? A review of studies on citing behavior. Journal of Documentation, 64(1):45- 80. DOI 10.1108/00220410810844150.", | |
| "links": null | |
| }, | |
| "BIBREF4": { | |
| "ref_id": "b4", | |
| "title": "Information Visualization: Beyond the Horizon", | |
| "authors": [ | |
| { | |
| "first": "Chaomei", | |
| "middle": [], | |
| "last": "Chen", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2006, | |
| "venue": "", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Chaomei Chen. 2006. Information Visualization: Be- yond the Horizon. Springer. 2nd Edition, Chapter 5.", | |
| "links": null | |
| }, | |
| "BIBREF5": { | |
| "ref_id": "b5", | |
| "title": "ParsCit: An open-source CRF reference string parsing package", | |
| "authors": [ | |
| { | |
| "first": "G", | |
| "middle": [], | |
| "last": "Isaac", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "C", | |
| "middle": [ | |
| "Lee" | |
| ], | |
| "last": "Councill", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Min-Yen", | |
| "middle": [], | |
| "last": "Giles", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "", | |
| "middle": [], | |
| "last": "Kan", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2008, | |
| "venue": "Proceedings of the Language Resources and Evaluation Conference (LREC-2008)", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Isaac G. Councill, C. Lee Giles, and Min-Yen Kan. 2008. ParsCit: An open-source CRF reference string parsing package. In Proceedings of the Language Resources and Evaluation Conference (LREC-2008), Marrakesh, Morocco.", | |
| "links": null | |
| }, | |
| "BIBREF6": { | |
| "ref_id": "b6", | |
| "title": "CiteWiz: A tool for the visualization of scientific citation networks", | |
| "authors": [ | |
| { | |
| "first": "Niklas", | |
| "middle": [], | |
| "last": "Elmqvist", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Philippas", | |
| "middle": [], | |
| "last": "Tsigas", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2004, | |
| "venue": "", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Niklas Elmqvist and Philippas Tsigas. 2004. CiteWiz: A tool for the visualization of scientific citation net- works. Technical Report CS:2004-05, Department of Computing Science, Chalmers University of Technol- ogy and G\u00f6teborg University, G\u00f6teborg, Sweden.", | |
| "links": null | |
| }, | |
| "BIBREF7": { | |
| "ref_id": "b7", | |
| "title": "Citation indexes for science: A new dimension in documentation through association of ideas", | |
| "authors": [ | |
| { | |
| "first": "Eugene", | |
| "middle": [], | |
| "last": "Garfield", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1955, | |
| "venue": "Science", | |
| "volume": "123", | |
| "issue": "", | |
| "pages": "108--111", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Eugene Garfield. 1955. Citation indexes for science: A new dimension in documentation through association of ideas. Science, 123:108-111.", | |
| "links": null | |
| }, | |
| "BIBREF8": { | |
| "ref_id": "b8", | |
| "title": "Can citation indexing be automated?", | |
| "authors": [ | |
| { | |
| "first": "Eugene", | |
| "middle": [], | |
| "last": "Garfield", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1965, | |
| "venue": "Statistical Association Methods for Mechanical Documentation. National Bureau of Standards", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Eugene Garfield. 1965. Can citation indexing be auto- mated? In Mary Elizabeth Stevens, Vincent E. Giu- liano, and Laurence B. Heilprin, editors, Statistical Association Methods for Mechanical Documentation. National Bureau of Standards, Washington, DC. NBS Misc. Pub. 269.", | |
| "links": null | |
| }, | |
| "BIBREF9": { | |
| "ref_id": "b9", | |
| "title": "Automated classification of citations using linguistic semantic grammars", | |
| "authors": [ | |
| { | |
| "first": "Mark", | |
| "middle": [], | |
| "last": "Garzone", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1996, | |
| "venue": "", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Mark Garzone. 1996. Automated classification of ci- tations using linguistic semantic grammars. Master's thesis, Dept. of Computer Science, The University of Western Ontario, Canada.", | |
| "links": null | |
| }, | |
| "BIBREF10": { | |
| "ref_id": "b10", | |
| "title": "Binary codes capable of correcting deletions, insertions, and reversals", | |
| "authors": [ | |
| { | |
| "first": "Vladimir", | |
| "middle": [ | |
| "I" | |
| ], | |
| "last": "Levenshtein", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1966, | |
| "venue": "Soviet Physics Doklady", | |
| "volume": "10", | |
| "issue": "8", | |
| "pages": "707--710", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Vladimir I. Levenshtein. 1966. Binary codes capable of correcting deletions, insertions, and reversals. Soviet Physics Doklady, 10(8):707-710.", | |
| "links": null | |
| }, | |
| "BIBREF11": { | |
| "ref_id": "b11", | |
| "title": "A design methodology for a biomedical literature indexing tool using the rhetoric of science", | |
| "authors": [ | |
| { | |
| "first": "Robert", | |
| "middle": [], | |
| "last": "Mercer", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Chrysanne Di", | |
| "middle": [], | |
| "last": "Marco", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2004, | |
| "venue": "Linking Biological Literature, Ontologies and Databases", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "77--84", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Robert. E Mercer and Chrysanne Di Marco. 2004. A design methodology for a biomedical literature index- ing tool using the rhetoric of science. In Lynette Hirschman and James Pustejovsky, editors, HLT- NAACL 2004 Workshop: BioLINK 2004, Linking Bi- ological Literature, Ontologies and Databases, pages 77-84, Boston, Massachusetts, USA.", | |
| "links": null | |
| }, | |
| "BIBREF12": { | |
| "ref_id": "b12", | |
| "title": "Five papers on WordNet", | |
| "authors": [ | |
| { | |
| "first": "George", | |
| "middle": [ | |
| "A" | |
| ], | |
| "last": "Miller", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Richard", | |
| "middle": [], | |
| "last": "Beckwith", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Christiane", | |
| "middle": [], | |
| "last": "Fellbaum", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Derek", | |
| "middle": [], | |
| "last": "Gross", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Katherine", | |
| "middle": [ | |
| "J" | |
| ], | |
| "last": "Miller", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 1993, | |
| "venue": "", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "George A. Miller, Richard Beckwith, Christiane Fell- baum, Derek Gross, and Katherine J. Miller. 1993. Five papers on WordNet. Technical report, Cognitive Science Laboratory, Princeton University.", | |
| "links": null | |
| }, | |
| "BIBREF13": { | |
| "ref_id": "b13", | |
| "title": "Visualization of relational structure among scientific articles", | |
| "authors": [ | |
| { | |
| "first": "Mao", | |
| "middle": [ | |
| "Lin" | |
| ], | |
| "last": "Quang Vinh Nguyen", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Simeon", | |
| "middle": [], | |
| "last": "Huang", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "", | |
| "middle": [], | |
| "last": "Simoff", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2007, | |
| "venue": "Advances in Visual Information Systems", | |
| "volume": "4781", | |
| "issue": "", | |
| "pages": "415--425", | |
| "other_ids": { | |
| "DOI": [ | |
| "10.1007/978-3-540-76414-440" | |
| ] | |
| }, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Quang Vinh Nguyen, Mao Lin Huang, and Simeon Simoff. 2007. Visualization of relational structure among scientific articles. Advances in Visual Informa- tion Systems, pages 415-425. Springer LNCS 4781, DOI 10.1007/978-3-540-76414-4 40.", | |
| "links": null | |
| }, | |
| "BIBREF14": { | |
| "ref_id": "b14", | |
| "title": "The use and misuse of journal metrics and other citation indicators. Archivum Immunologiae et Therapiae Experimentalis", | |
| "authors": [ | |
| { | |
| "first": "A", | |
| "middle": [], | |
| "last": "David", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "", | |
| "middle": [], | |
| "last": "Pendlebury", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2009, | |
| "venue": "", | |
| "volume": "57", | |
| "issue": "", | |
| "pages": "1--11", | |
| "other_ids": { | |
| "DOI": [ | |
| "10.1007/s00005-009-0008-y" | |
| ] | |
| }, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "David A. Pendlebury. 2009. The use and misuse of jour- nal metrics and other citation indicators. Archivum Im- munologiae et Therapiae Experimentalis, 57(1):1-11. DOI 10.1007/s00005-009-0008-y.", | |
| "links": null | |
| }, | |
| "BIBREF15": { | |
| "ref_id": "b15", | |
| "title": "The ACL anthology network corpus", | |
| "authors": [ | |
| { | |
| "first": "R", | |
| "middle": [], | |
| "last": "Dragomir", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Pradeep", | |
| "middle": [], | |
| "last": "Radev", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Vahed", | |
| "middle": [], | |
| "last": "Muthukrishnan", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "", | |
| "middle": [], | |
| "last": "Qazvinian", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2009, | |
| "venue": "Proceedings of the ACL Workshop on Natural Language Processing and Information Retrieval for Digital Libraries", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Dragomir R. Radev, Pradeep Muthukrishnan, and Vahed Qazvinian. 2009. The ACL anthology network cor- pus. In Proceedings of the ACL Workshop on Nat- ural Language Processing and Information Retrieval for Digital Libraries, Singapore.", | |
| "links": null | |
| }, | |
| "BIBREF16": { | |
| "ref_id": "b16", | |
| "title": "Extracting and querying relations in scientific papers", | |
| "authors": [ | |
| { | |
| "first": "Ulrich", | |
| "middle": [], | |
| "last": "Sch\u00e4fer", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Hans", | |
| "middle": [], | |
| "last": "Uszkoreit", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Christian", | |
| "middle": [], | |
| "last": "Federmann", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Torsten", | |
| "middle": [], | |
| "last": "Marek", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Yajing", | |
| "middle": [], | |
| "last": "Zhang", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2008, | |
| "venue": "Proceedings of the 31st Annual German Conference on Artificial Intelligence", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "127--134", | |
| "other_ids": { | |
| "DOI": [ | |
| "LNAI5243.DOI10.1007/978-3-540-85845-416" | |
| ] | |
| }, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Ulrich Sch\u00e4fer, Hans Uszkoreit, Christian Federmann, Torsten Marek, and Yajing Zhang. 2008. Extract- ing and querying relations in scientific papers. In Proceedings of the 31st Annual German Conference on Artificial Intelligence, KI 2008, pages 127-134, Kaiserslautern, Germany. Springer LNAI 5243. DOI 10.1007/978-3-540-85845-4 16.", | |
| "links": null | |
| }, | |
| "BIBREF17": { | |
| "ref_id": "b17", | |
| "title": "Automatic classification of citation function", | |
| "authors": [ | |
| { | |
| "first": "Simone", | |
| "middle": [], | |
| "last": "Teufel", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Advaith", | |
| "middle": [], | |
| "last": "Siddharthan", | |
| "suffix": "" | |
| }, | |
| { | |
| "first": "Dan", | |
| "middle": [], | |
| "last": "Tidhar", | |
| "suffix": "" | |
| } | |
| ], | |
| "year": 2006, | |
| "venue": "Proceedings of the 2006 Conference on Empirical Methods in Natural Language Processing", | |
| "volume": "", | |
| "issue": "", | |
| "pages": "103--110", | |
| "other_ids": {}, | |
| "num": null, | |
| "urls": [], | |
| "raw_text": "Simone Teufel, Advaith Siddharthan, and Dan Tidhar. 2006. Automatic classification of citation function. In Proceedings of the 2006 Conference on Empirical Methods in Natural Language Processing, pages 103- 110, Sydney, Australia.", | |
| "links": null | |
| } | |
| }, | |
| "ref_entries": { | |
| "FIGREF0": { | |
| "text": "Workflow from ACL Anthology data (top) to citation graph navigation applet and citation sentence viewer (bottom)", | |
| "num": null, | |
| "type_str": "figure", | |
| "uris": null | |
| }, | |
| "FIGREF1": { | |
| "text": "Typed citation graph navigator applet", | |
| "num": null, | |
| "type_str": "figure", | |
| "uris": null | |
| }, | |
| "FIGREF2": { | |
| "text": "Focused paper in the center", | |
| "num": null, | |
| "type_str": "figure", | |
| "uris": null | |
| }, | |
| "FIGREF3": { | |
| "text": "Papers cited by the focused paper", | |
| "num": null, | |
| "type_str": "figure", | |
| "uris": null | |
| }, | |
| "FIGREF4": { | |
| "text": "Papers citing the focused paper", | |
| "num": null, | |
| "type_str": "figure", | |
| "uris": null | |
| }, | |
| "FIGREF5": { | |
| "text": "Fan-out layout: avoid crossing edges caused by citations on the same level", | |
| "num": null, | |
| "type_str": "figure", | |
| "uris": null | |
| }, | |
| "FIGREF6": { | |
| "text": "Order: avoid crossing edges by ordering chronologically (strict, simple variant on the left for comparison), taking into account the position of the cited paper on the previous level (right)", | |
| "num": null, | |
| "type_str": "figure", | |
| "uris": null | |
| }, | |
| "TABREF0": { | |
| "content": "<table><tr><td>Classes</td><td colspan=\"2\">Citations Percent</td></tr><tr><td>Agree</td><td>3513</td><td>3.8%</td></tr><tr><td>Agree, Neutral</td><td>2020</td><td>2.2%</td></tr><tr><td>Negative</td><td>1147</td><td>1.2%</td></tr><tr><td>PRecycle</td><td>10609</td><td>11.6%</td></tr><tr><td>PRecycle, Agree</td><td>1419</td><td>1.6%</td></tr><tr><td>PRecycle, Agree, Neutral</td><td>922</td><td>1.0%</td></tr><tr><td>PRecycle, Neutral</td><td>3882</td><td>4.2%</td></tr><tr><td>Neutral</td><td>13430</td><td>14.7%</td></tr><tr><td>Undef</td><td>54837</td><td>60.0%</td></tr></table>", | |
| "text": "", | |
| "html": null, | |
| "num": null, | |
| "type_str": "table" | |
| }, | |
| "TABREF1": { | |
| "content": "<table/>", | |
| "text": "", | |
| "html": null, | |
| "num": null, | |
| "type_str": "table" | |
| } | |
| } | |
| } | |
| } |