cs0211041/Draft.tex
1: \documentstyle[12pt,epsf]{article}
2: 
3: \setlength{\textwidth}{160mm}
4: \setlength{\textheight}{220mm}
5: \setlength{\topmargin}{0mm}
6: \setlength{\oddsidemargin}{0mm}
7: \setlength{\evensidemargin}{0mm}
8: 
9: \makeatletter
10: \def\lesssim{\mathrel{\mathpalette\vereq<}}
11: \def\vereq#1#2{\lower3pt\vbox{\baselineskip1.5pt \lineskip1.5pt
12: \ialign{$\m@th#1\hfill##\hfil$\crcr#2\crcr\sim\crcr}}}
13: \def\gtrsim{\mathrel{\mathpalette\vereq>}}
14: \def\alt{\lesssim}
15: \def\agt{\gtrsim}
16: \makeatother
17: 
18: \title{
19: \begin{flushright}\normalsize
20: \vspace{0cm}
21: Internal Report~~\\
22: DESY L-02-02~~~\\
23: November 2002
24: \vspace{0.cm}
25: \end{flushright}
26: \begin{center}
27: \centerline{\large\bf An Approach to Automatic Indexing of Scientific
28: Publications}
29: \vspace*{2mm}
30: \centerline{\large\bf in High Energy Physics for Database SPIRES HEP}  
31: \end{center}
32: }
33: %\title{An Approach to Automatic Indexing of Scientific
34: %Publications in High Energy Physics for Database SPIRES-HEP}
35: 
36: \author{A.V.~Averin$^{a)}$, L.A.~Vassilevskaya$^{b)}$\\
37:        {\small\it a) NSI, B.Tulskaya 52, 113191 Moscow, Russia} \\
38:        {\small\it (E-mail address: ava@ibrae.ac.ru)} \\
39:        {\small\it b) Deutsches Elektronen-Synchrotron DESY, D-22603
40:        Hamburg, Germany} \\
41:        {\small\it (E-mail address: lyuba@mail.desy.de)}  }
42: 
43: \date{}
44: 
45: \begin{document}
46: 
47: \maketitle
48: 
49: \begin{abstract}
50: \noindent
51: %
52: %!!!
53: %We introduce an approach to automatic content-based
54: We introduce an approach to automatic 
55: %
56: indexing of e-prints based on a
57: pattern-matching technique making extensive use of an
58: Associative Patterns Dictionary (APD), developed by us. Entries in
59: the APD consist of natural language phrases with the same semantic
60: interpretation as a set of keywords from a controlled vocabulary.
61: The method also allows to recognize within e-prints formulae
62: written in \TeX{} notations that might also appear as keywords.
63: We present an automatic indexing system, AUTEX, which we have applied
64: to keyword index e-prints in selected areas in high energy physics
65: (HEP) making use of the DESY-HEPI thesaurus as a controlled vocabulary.
66: \end {abstract}
67: 
68: \section{Introduction}
69: \label{Sec:introduction}
70: 
71: Indexing is considered classically as an evaluation of documents,
72: leading to its content-based representation, which could subsequently
73: be accessed using a descriptor-based information retrieval system.
74: These descriptors (keywords) are taken as a rule from a well-defined
75: and controlled vocabulary. One expects that this controlled vocabulary
76: is periodically updated to be {\it en vogue}, i.e., it would reflect new
77: fields, and new trends and innovations within the established fields.\\
78: 
79: \noindent
80: Intellectual indexing implies at least two essential
81: steps: the conceptual grasp of the content of a
82: document (content analysis), and transcribing the content in terms of
83: index descriptors (taken, for example, from a thesaurus). The central
84: point is that the intellectual indexing process requires 
85:  an understanding by a human indexer of the
86: ``aboutness'' \cite{Chung} of a document being indexed.
87: Historically, this was the only method of indexing the literature,
88: as all publishing and related bibliographical work was done with
89: papers. This classic form of document analysis is still widely used,
90: and is particularly important in the circumstances where the texts
91: are available in a non-electronic form or for non-text documents
92: (pictures, photographs, films etc.). The down side of the
93: intellectual method is that it is expensive, subject-oriented, and hence 
94: localized, and time consuming. Moreover, human-based judgment errors can 
95: not always be avoided.\\
96: 
97: \noindent
98:  In the case of High Energy
99: Physics (HEP), which is the primary area of research we shall
100: concentrate on, keyword indexing has been done at the German  High
101: Energy Physics
102: Laboratory DESY (Deutsches Elektronen-Synchrotron)~\cite{DESY} 
103: in Hamburg, which has developed and maintained the High Energy 
104: Physics Index (HEPI) since 1963.
105: The DESY-HEPI keywords~\cite{HEPI} are included in all records
106: (published papers, preprints, conference proceedings and books) in the
107: HEP literature database called SPIRES (Stanford Public Information REtrieval System),
108: maintained at the high energy physics laboratory SLAC (Stanford Linear
109: Accelerator Center) in California~\cite{SPIRES}. This keyword
110: indexing has been done exclusively using the so-called
111: intellectual method in the Scientific Documentation Group at DESY,
112: involving a team which includes some members with HEP background.
113: The thesaurus used in HEPI is based on approximately 2500 terms, and
114: is occasionally updated. With the growing use of computer-based methods
115: and the rapid  increase in the  number of scientific publications, most of
116: which are prepared in electronic form, and a well-maintained central
117: repository of e-prints in HEP and related fields now based at Cornell
118: University~\cite{Ginsparg}, there is an  overriding need to develop 
119: methods of automatic
120: indexing. In fact, automatic indexing is in the long term {\it the only
121: realistic and economically viable method} of keyword indexing
122: scientific documents.\\
123: 
124: \noindent
125: However, despite all the technological
126: progress and, what concerns HEP and several related fields, the
127: availability of e-prints from the electronic  archives dating
128: back to the pioneering work of Ginsparg done at the Los~Alamos National
129: Laboratory in New Mexico in 1991~\cite{Ginsparg}, there is no
130: fully automatic indexing system in HEP available yet. In the
131: recent past, some progress in this direction has been made at
132: CERN (the European Laboratory for Particle Physics) in Geneva.
133: The first of these approaches called SOCRATES  was an attempt to
134: incorporate software for automatic indexing \cite{Dallman/Meur},
135: based on constructing a set of the longest noun-type phrases from
136: the abstracts of documents. This approach was, however, abandoned
137: at the end of 2000 due to the difficulties of integrating its
138: software into the database system of the CERN library. This was
139: followed by a second research and development (R\&D) project
140: called HEPindexer~\cite{Raez/Dallman}, in which the first step aimed 
141: at the generation of the DESY-HEPI keywords has been implemented 
142: following a statistical method. The results of HEPindexer,
143: which was trained on some 2400 high energy physics papers and tested
144: so far on some 1200 additional documents, is quite impressive in
145: that the results were close to 60\% in precision and recall
146: benchmarked against the HEPI-SPIRES method of keyword
147: indexing~\cite{SPIRES}. However, HEPindexer is not yet in use for 
148: professional indexing of the HEP documents.\\
149: 
150: \noindent
151: In this report we describe our approach to automatic indexing of HEP
152: e-prints based on {\it associative pattern} matching. Central to 
153: our approach is the {\it Associative Patterns Dictionary}, which we
154: have  developed and encoded.
155: Our software package, AUTEX (AUTomatic indEXer) has been tested so
156: far in a limited area of HEP, namely neutrino physics, astrophysics
157: and physics of the axion. This choice is  based on our expertise and
158: HEP background. The goal of AUTEX indexer is to take a document (an 
159: e-print) as input and represent its contents as a set of keywords 
160: from the DESY-HEPI thesaurus. 
161: %It makes use of English and \TeX{} grammar. 
162: Judging from the number of e-prints in these fields, our project
163: covers at present about 15\% of the e-prints in HEP. This is also
164: reflected in the number of the keywords presently encoded in the AUTEX
165: database (approximately 550) compared to approximately 2500 of the
166: DESY-HEPI thesaurus. We have tested our software package on about
167: 300 e-prints in these fields and have compared the results of our
168: automatic indexing with the ones generated using the SPIRES-HEP literature
169: database. Our system can be accessed also by remote users and is 
170: already in a position to be used for indexing work in the stated 
171: areas of HEP.  We describe the 
172: salient features of our automatic indexer AUTEX, showing some 
173: representative results of the index reports on e-prints together 
174: with comparisons
175: with the SPIRES HEP generated intellectual indexing results. Templates
176: of the AUTEX screen at various intermediate steps 
177: of automatic indexing are also shown. In future, we plan to enlarge
178: the applications to cover the entire fields of HEP, astrophysics and 
179: cosmology, and to include additional features in the indexing 
180: reports such as the titles of the papers, authors names, and their 
181: institutional affiliations. Further details and AUTEX report 
182: results can be seen in Ref.~\cite{Vassilevskaya}. 
183: 
184: \section{AUTEX - an Automatic Indexer \\
185: of High Energy Physics Documents}
186: \label{Sec:system}
187: \vspace{5mm}
188: 
189: We start by describing the basic components of the AUTEX system.
190: These include the terminology, the {\it Associative Patterns 
191: Dictionary}, an introduction to the basic principle on which
192: AUTEX is based, and the user interface. 
193: 
194: \subsection{Terminology}
195: To begin, one should outline some definitions.\\
196: 
197: {\large\bf Keywords List} \\
198: 
199: \noindent
200: {\it Keywords List} is a subset of a controlled vocabulary.
201: In the case of high energy physics it is assumed to be covered by
202: DESY-HEPI thesaurus.  A {\it keyword} can be either a separate word: \\
203: \hspace*{5,0mm} {\bf astrophysics} \\
204: \hspace*{5,0mm} {\bf neutrino} \\
205: or a combination of some words: \\
206: \hspace*{5,0mm} {\bf dispersion relations} \\
207: \hspace*{5,0mm} {\bf interpretation of experiments}.\\
208: 
209: {\large\bf Keychain} \\
210: 
211: \noindent
212: {\it Keychain} is composed of keywords written in one line and joined
213: by a comma and a blank as a delimiter:
214: %
215:  \[
216: \underbrace{\overbrace{\bf astrophysics}^{keyword}}_{keychain}
217: \]
218: %
219: \[
220: \underbrace{
221: \overbrace{\bf neutrino}^{keyword}, \,
222: \overbrace{\bf magnetic \,\, moment}^{keyword}
223: }_{keychain}
224: \]
225: %
226: \[
227: \underbrace{
228: \overbrace{\bf plasmon}^{keyword}, \,
229: \overbrace{\bf longitudinal}^{keyword}, \,
230: \overbrace{\bf dispersion \, \, relations}^{keyword} \,
231: }_{keychain}
232: \]
233: %
234: \[
235: \dots
236: \]
237: 
238: \noindent
239: In principle the
240: system is able to construct a keychain of any length (any
241: number of keywords). The keychain of length one is just
242: a keyword.
243: At the moment, keychains of length no more than two are used
244: {\it de facto} within the Scientific Documentation Group at DESY. It
245: is keychains -- not keywords -- that are assigned to a document during
246: indexing process. The resulting set of keychains forms a document
247: index that is incorporated into the SPIRES-HEP database. \\
248: 
249: 
250: {\large\bf Associative Patterns Dictionary (APD)} \\
251: 
252: \noindent
253: In its simplest	form an {\bf associative pattern} is thought of as any
254: English phrase that might also include a formula written in \TeX{}
255: notations:\\
256: \hspace*{5,0mm}	{\tt astrophysics} \\
257: \hspace*{5,0mm}	{\tt neutron stars} \\
258: \hspace*{5,0mm}	{\tt search for galactic dark matter} \\
259: \hspace*{5,0mm}	{\tt \verb|$\nu \to \nu \gamma$|} \\
260: \hspace*{5,0mm}	{\tt axion decay \verb|$\a \to e^+ e^-$|.} \\
261: 
262: \noindent
263: Given an associative pattern, we assign to it a set of keychains
264: defining the {\it meaning} of this pattern. That is, one gets a
265: semantic interpretation of the pattern in terms of keychains from
266: the controlled vocabulary. Applying an alternation metasymbol {\bf $|$}
267: (vertical bar) makes it possible to consider multiple associative
268: patterns of the form:
269: 
270: \begin{center}
271: {\tt energy dissipation $|$ dissipation of energy $|$ energy is 
272: dissipated}.\\ 
273: \end{center}
274: 
275: \noindent
276: In most cases each
277: constituent phrase in an associative pattern is a noun phrase.
278: However, in principle,  {\it any} other phrase is allowed.
279: Here are some examples of how the APD entrance may look like: 
280: 
281: \begin{center}
282: \begin{tabular}{llr}
283: %
284: $\fbox{\parbox{3.5cm}{
285: {\tt leptogenesis} 
286: }}$ 
287: & $\Longrightarrow$ 
288: $\fbox{\parbox{4cm}{
289: {\bf lepton, production} 
290: }}$ \\
291: 
292: \vspace{1mm} & \\
293: 
294: %
295: $\fbox{\parbox{5.8cm}{
296: {\tt abelian horizontal charge} {\bf $|$} \\
297: {\tt horizontal abelian charge}
298: }}$ 
299: & $\Longrightarrow$ 
300: $\fbox{\parbox{4.5cm}{
301: {\bf horizontal symmetry} \\
302: {\bf charge, abelian}
303: }}$ \\
304: 
305: \vspace{1mm} & \\
306: 
307: %
308: $\fbox{\parbox{6cm}{
309: {\tt axion decay into electron- positron pair} {\bf $|$} 
310: {\tt axion decay $a \to e^+ e^-$} {\bf $|$} 
311: {\tt electron-positron pair production by axion} 
312: }}$ 
313: & $\Longrightarrow$ 
314: $\fbox{\parbox{6cm}{
315: {\bf axion, leptonic decay} \\
316: {\bf electron, pair production} \\
317: {\bf axion $\to$ positron electron} 
318: }}$ \\
319: 
320: \vspace{1mm} & \\
321: %
322: $\fbox{\parbox{6.5cm}{
323: {\tt neutrino pair production by a virtual photon}  {\bf $|$}
324: $\gamma_{virt} \to \nu \bar \nu$ 
325: }}$ 
326: & $\Longrightarrow$ 
327: $\fbox{\parbox{6,7cm}{
328: {\bf neutrino, pair production} \\
329: {\bf neutrino, photoproduction} \\
330: {\bf photon, off-shell} \\
331: {\bf photon $\to$ neutrino antineutrino} 
332: }}$ 
333: %
334: \end{tabular}
335: \end{center}
336: 
337: \noindent
338: Therefore, each entrance in the APD represents a relation, i.e., an
339: {\it association} between an {\it idea} and a set of index terms.
340: 
341: \subsection{How does AUTEX work?}
342: 
343: As already stated, the AUTEX indexer takes a document (an e-print) as
344: input and represents its content as a set of keywords from DESY-HEPI
345: thesaurus. Of course, this thesaurus is a default option currently and
346: as the applications increase, the DESY-HEPI can be replaced by another 
347: thesaurus.
348: Rather than thinking of a document as natural language input text
349: (which is additionally wrapped by \TeX{} commands), we treat
350: it as a stream of characters and apply the
351: pattern-matching technique. At the present stage, the AUTEX
352: system does not perform aspects of natural language
353: processing which traditionally include part-of-speech
354: tagging and syntactic analysis. Instead, each associative
355: pattern from the APD tries to match a document content.
356: If it succeeds, the corresponding set of keychains, associated
357: with this pattern, is added into a document index.
358: So, the AUTEX indexer does not distinguish whether a phrase
359: is a noun phrase or it is something else. It just matches
360: patterns. \\
361: 
362: \noindent
363: The indexer actually consists of two phases of operation: {\it
364: preprocessing} and {\it pattern matching}. The first
365: one in our case involves undertaking some {\it preprocessing} of the input
366: text. The system provides a set of built-in {\it pointers} that
367: enable us to specify which parts of the article to work out.
368: Currently, the following six pointers are available: ``title'',
369: ``abstract'', ``caption'', ``section'', ``conclusions'', and
370: ``full-text''.
371: %
372: Once a pointer is set, the system parser
373: extracts the corresponding part of the document and includes it
374: into the text to be processed. These parts of the text usually
375: (though not always) contain an important information on the
376: purpose of investigations and/or the results obtained in the
377: paper. So, reading these parts of the article is expected to lead to
378: information on the text relevant for keyword indexing.
379: Of course, multiple choice is
380: possible and is used in the actual indexing work presented here.
381: Note also that the ``section'' pointer
382: instructs the parser to fetch titles of the various sections (not the
383: section contents themselves) including the titles of ``subsections'',
384: ``subsubsections'', etc. as well. \\
385: 
386: \noindent
387: Another goal of the preprocessing is to prune irrelevant
388: \TeX{} commands from each part to be pointed out. For example,
389: the pattern {\tt strong magnetic field} has to match the occurrence
390: {\tt \verb|{\it strong}| magnetic field} as well. \\
391: 
392: \subsection{User Interface}
393: 
394: \noindent
395: For the AUTEX indexer to work it has to be supplied with an
396: appropriate look up dictionary -- APD in our case. It is obvious, that
397: the APD is of crucial importance for an effective and correct 
398: automatic indexing. In particular, how the  APD is populated and 
399: maintained will play a central role in indexing.
400: At the current stage
401: of development of the AUTEX project, APD is populated
402: ``by hand''. This work is done by a subject specialist
403: making use of a graphic user interface provided by the system.
404: This interface enables navigation across the system and
405: facilitates the management of Keyword List, Keychain
406: List, APD, and the indexer itself. For example, within APD Manager,
407: the APD records can be accessed using a flexible filter.
408: The filter options include keying of a truncated string
409: of characters, alphabetical selection, and keychain
410: selection. This enables one to select all patterns associated
411: with a given set of keychains. Similar filters are
412: provided in Keyword Manager and Keychain Manager as well.
413: 
414: \section{Pattern Matching}
415: 
416: \noindent
417: The patterns used in AUTEX pattern matching are, in fact,
418: regular expressions in their simplest form. Let us suppose that we are 
419: creating a pattern that will succeed if an input text is about massless 
420: neutrino. It may be seen, for example,
421: that the pattern {\tt massless~neutrino} will not match the
422: occurrence {\tt massless~chiral~neutrinos}. Such cases are easily
423: resolved if one introduces the following regular expression 
424: 
425: \begin{center}
426: {\tt \verb|massless[ \w]+neutrinos?|} \\
427: \end{center}
428: 
429: \noindent
430: where the metacharacter {\sf [\,\,]} and the
431: quantifiers {\sf ?} and {\sf +} have their standard 
432: meaning in the extended regular expression syntax that is supported by 
433: {\sf egrep} routine.
434: The backslashed metacharacter {\tt \verb|\w|} matches a single
435: alphanumeric character.
436: %
437: This pattern will succeed whatever words occur between the words
438: {\tt massless} and {\tt neutrino}, but it will still fail on the
439: phrase, for example, {\tt the neutrino is assumed to be massless}.
440: To match this phrase too, we modify our pattern to yield
441: 
442: \begin{center}
443: {\tt \verb|massless[ \w]+neutrinos?|{\bf $|$}\verb|neutrinos?[
444: \w]+massless|} \\
445: \end{center}
446: 
447: \noindent
448: This improves pattern matching. However, it will occasionally also process
449: English text in which a certain pattern is matched though the sense in
450: which these words are used is different than what a certain {\it 
451: keychain} implies. For example, 
452: one may encounter the following text in a document:
453: {\tt neutrino decays into some particles one of which
454: has to be massless}.
455: The last pattern given above will match this text. However,
456: this will be a wrong matching in the sense of the content, as the text 
457: does not discuss  massless neutrino but rather a massless particle which 
458: is the decay product of a neutrino. \\
459: 
460: \noindent
461: Therefore, if we want that our pattern does not miss any relevant
462: sentence about massless neutrino and all matches correctly depict the
463: content of a document, then one has to add further discriminatory
464: characteristics in the pattern matching engine.
465: In particular, this engine has to be able to first understand the
466: grammar of the sentence. Regular expressions are extremely powerful
467: tool.
468: However, they are not sufficient for grammar understanding. So, we must
469: integrate into the system lexical and syntactic analyzers of
470: English.  The idea is to identify in the original text some
471: constituents so that they will be recognized by a limited set of
472: associative patterns, filtering out peripheral text portions. 
473: There is an additional complication in our case due to the \TeX{} 
474: formulae. This requires further development work and 
475: we hope to return to this problem in a future publication.
476: \\
477: 
478: \noindent
479: Note that we stay within the framework based on an human-made
480: APD rather than on any statistical model. That is, each associative
481: pattern is built by a human -- a subject specialist -- and each
482: association is assigned by an expert. This approach results in rather
483: accurate and reliable index. Our first experiments show that, at
484: this stage, only a few index terms are lost.
485: The system can be sufficiently trained eventually to reduce this loss to
486: a bare minimum. We have striven at reducing the appearance of
487: redundant or irrelevant terms in the index, and our studies show that
488: in this we have largely succeeded.\\
489: 
490: \section{Results and Concluding Remarks}
491: 
492: \noindent
493: We conclude in this section by  showing some representative results
494: from the AUTEX indexer. The results of keyword indexing ten 
495: documents from the arXiv.org e-print archive are shown in Appendix 
496: A. The indexing reports from the AUTEX indexer are compared with the
497: already existing SPIRES HEP reports, both of which use the DESY-HEPI
498: thesaurus. The papers discuss mostly neutrino physics, and in one case
499: physics of the axion.
500: Many more such comparisons can be seen in Ref.~\cite{Vassilevskaya}. These 
501: comparisons show that the AUTEX indexer does a comprehensive 
502: job in indexing e-prints, and some of the keywords in the AUTEX 
503: reports are missing from the SPIRES HEP generated reports. Certainly, 
504: there is some room for improvement in the AUTEX system as well, but 
505: the efficiency already reached in the areas covered by AUTEX is very 
506: high.\\
507: 
508: \noindent
509: We show some templates of the AUTEX screen at important stages 
510: of automatic indexing.
511: Figs.~1 and 2 show the examples
512: of the user interface screens corresponding to the two Managers
513: called {\it Keyword Manager} and {\it Keychain Manager}, 
514: respectively. 
515: The {\it Article Manager}, whose template from the AUTEX screen is shown 
516: in Fig.~3, enables one
517: to upload an article file, set up the 
518: pointers mentioned in section 2, submit selected articles in 
519: batch mode, and control the resulting index reports.
520: The AUTEX report includes the article index along with the 
521: information which indicates from which parts of the article
522: (title, abstract, etc.) each index term stems. A typical 
523: indexing result is shown in  Fig.~4. \\
524: 
525: \noindent
526: Within the AUTEX project, the high 
527: priority for us is to enlarge the
528: applications (done by enlarging the dictionary APD) to cover the
529: entire field of high energy physics, cosmology and astrophysics. 
530: This is a straight forward though time-consuming step. However, the 
531: results achieved so far are very encouraging and we trust that the goal of
532: developing a completely automatic system of indexing and retrieval of 
533: all e-prints in the fields of HEP, astrophysics and cosmology can be 
534: achieved. Already at its present stage AUTEX is a useful tool, and it can 
535: be used to help in indexing e-prints in some areas of HEP and 
536: astrophysics.\\
537:  
538: \noindent {\bf Acknowledgment}
539: L.V. would like to thank Dietmar Schmidt for his support and
540: encouragement during the course of this work. We thank David Dallman for
541: reading the manuscript. 
542: 
543: \newpage
544: \setcounter{figure}{0}
545: \begin{figure} %\label{fig:keyword}
546: \begin{center}
547: \epsfxsize=1.1\textwidth
548: \epsffile{keyword.eps}
549: %
550: \caption{ Base screen of the AUTEX Keyword Manager. Keywords are 
551: browsed in the most convenient way as a filtered list (shown on the 
552: left). The filter has two options: 1) one can choose a 
553: letter from the alphabet popup menu  for the system to select all 
554: keywords starting from this letter (here, the letter L), or 
555: 2) one can specify a truncated string (more than one character). 
556: In this case, the system will return the list of keywords with these 
557: characters as leading. If both filter options are set, the list of 
558: keywords will satisfy both of them (logical AND operation). 
559: }
560: \end{center}
561: \label{Fig1}
562: \end{figure}
563: 
564: \setcounter{figure}{1}
565: \begin{figure} %\label{chain}
566: \begin{center}
567: \epsfxsize=1.1\textwidth
568: \epsffile{chain.eps}
569: % 
570: \caption{A template from the AUTEX screen showing
571: %
572: %!!!
573: %some {\it keychains}. 
574: a keychain creation/modification.
575: %
576: One can compose a keychain of any length  (any number of
577: keywords)  and change the order of keywords within the keychain. To
578: add a keyword  into a chain, one has to just press this keyword from
579: the filtered list. This keyword immediately appears in the chain.
580: }
581: \end{center}
582: \label{Fig2}
583: %\label{chain}
584: \end{figure}
585: 
586: \setcounter{figure}{2}
587: \begin{figure} %\label{queue}
588: \begin{center}
589: \epsfxsize=1.1\textwidth
590: \epsffile{queue.eps}
591: % 
592: \caption{A template from the AUTEX screen, showing a typical setup
593: with multiple pointers
594: %
595: %!!!
596: %, references to the e-Prints
597: %which can be download sequentially  forming the 
598: %process queue which then can be indexed in the batch submission 
599: %mode.
600: along with the process queue, that is a list of references to the
601: e-prints that have already been uploaded into the system, but not
602: indexed yet. The e-prints from the process queue then can be indexed 
603: in the batch submission mode.
604: %
605: }
606: \end{center}
607: \label{Fig3}
608: %\label{Fig:3}
609: \end{figure}
610: 
611: \setcounter{figure}{3}
612: \begin{figure} %\label{report}
613: \begin{center}
614: \epsfxsize=1.1\textwidth
615: \epsffile{report.eps}
616: % 
617: \caption{A template from the AUTEX screen, showing the results of 
618: the indexed e-print (here hep-ph/0106157). 
619: The results can be 
620: previewed, corrected, printed, and stored in a file. One can edit an 
621: article profile that includes, at present, the SLAC ID and prefix.
622: In future, article title, authors names and addresses are planed to 
623: be incorporated into the profile.} 
624: \end{center}
625: \label{Fig4}
626: \end{figure}
627: %\\
628: 
629: \newpage
630: \addcontentsline{toc}{section}{References}
631: \begin{thebibliography}{100}
632: \label{Sec:bibliography}
633: 
634: \bibitem[Chung 1998]{Chung}
635: Yi-Ming Chung, William M. Pottenger, Bruce R. Schatz:\\
636: Automatic Subject Indexing Using an Associative Neural Network\\
637: (URL:http://www.cannis.uiuc.edu/chung98automatic.html).
638: 
639: \bibitem[Dallman/Meur 1999]{Dallman/Meur}
640: Dallman, D., Le Meur, J.Y.:
641: \newblock {Automatic Keywording of High Energy Physics.}
642: \newblock In 4th International Conference on Grey Literature: New
643: Frontiers in Grey Literature, Washington, DC, USA, 1999.
644: 
645: \bibitem[DESY]{DESY}
646: Deutsches~Elektronen-Synchrotron~DESY,~Hamburg \\
647: (URL: http://www.desy.de/).
648: 
649: \bibitem[Ginsparg 1991]{Ginsparg}
650: Ginsparg, P.:
651: \newblock {arXiv.org e-print archive (URL:
652: http://www.\-arXiv.\-org)}.
653: \newblock {The e-print archive arXiv.org was initially based at Los Alamos 
654: National Laboratory.
655: It has followed its creator, Paul Ginsparg, to Cornell University in
656: Ithaca, New York,  last year, but it is still often referred to as the 
657: LANL archive}.
658: 
659: \bibitem[HEPI]{HEPI}
660: The HIGH ENERGY PHYSICS INDEX
661: \newblock {Keywords 1996/1997}
662: \newblock Internal Report, DESY L-96-01, January 1996.
663: 
664: \bibitem[R{\'a}ez/Dallman 2001]{Raez/Dallman}
665: R\'aez, a.M., Dallman, D.:
666: \newblock {Experiences in Automatic Keywording of Particle Physics
667: Literature.}
668: \newblock High Energy Physics Libraries Webzine, Issue 5, 2001
669: (URL: http://library.cern.ch/HEPLW/5/papers/3/).
670: 
671: \bibitem[SPIRES]{SPIRES}
672: Stanford Linear Accelerator Center
673: \newblock SLAC Library, SLAC SPIRES HEP Datebase
674: (URL: http://www.slac.stanford.edu/spires/hep).
675: 
676: \bibitem[Vassilevskaya 2002]{Vassilevskaya}
677: Vassilevskaya, L.~A.; 
678: An Approach to Automatic Indexing of Scientific Publications in High 
679: Energy Physics for Database SPIRES-HEP;
680: %{\it Abschlu\ss arbeit im 
681: %Lehrgebiet Inhaltliche Erschlie\ss ung; Institut f\"ur Information 
682: %und Dokumentation, Potsdam, FRG}. 
683: DESY Report DESY L-02-01 , November 2002.
684: 
685: 
686: \end{thebibliography}
687: 
688: \newpage
689: \appendix
690: \section{Appendix: AUTEX Reports (Examples)} 
691: \label{app:Glossary}
692: 
693: In this Appendix, we present some representative reports on 
694: subject keyword indexing generated by  
695: our automatic indexing system AUTEX and compare them with the 
696: already existing SPIRES reports. In all these examples, the papers refer 
697: to their E-Print archive numbers. The symbol ``0'' at the end of some 
698: reports means that this particular {\it keychain} has no relation to the 
699: paper. The {\it keychains} written below the horizontal line in some 
700: reports indicate mismatching {\it keychains} between the AUTEX and SPIRES
701: reports. 
702: 
703: \subsection{\, hep-ph/0106157}
704: \label{app:hep-0106157}
705: 
706: \begin{center}
707: {\small\bf Title:} 
708: {\bf GUT implications from neutrino mass} \\
709: \end{center}
710: 
711: \begin{abstract}
712: An overview is given of the experimental neutrino mixing results and types of 
713: neutrino models proposed, with special attention to the general features of 
714: various GUT models involving intra-family symmetries and horizontal flavor 
715: symmetries.  Many of the features are then illustrated by a specific 
716: $SO(10)$ SUSY GUT model formulated by S.M. Barr and the author which can 
717: explain all four types of solar neutrino mixing solutions by various choices 
718: of the right-handed Majorana mass matrix.  The quantitative nature of the 
719: model's large mixing angle solution is used to compare the reaches of a 
720: neutrino super beam and a neutrino factory for determining the small 
721: $U_{e3}$ mixing matrix element. \\
722: \end{abstract}
723: 
724: \begin{tabular}{lll}
725: \noindent
726: \underline{{\rm AUTEX Report:}} &  & \underline{{\rm SPIRES Report:}} \\ \\
727: %
728:  {\bf neutrino, mass  } & & {\tt neutrino, mass  } \\
729:  {\bf grand unified theory, SO(10) } & & 
730:  {\tt grand unified theory, SO(10) } \\
731:  {\bf symmetry, $U(1) \times Z_2 \times Z_2$ } & & 
732:          {\tt symmetry, $U(1) \times Z_2 \times Z_2$ } \\ 
733:  {\bf supersymmetry  } & & {\tt supersymmetry} \\
734:  {\bf horizontal symmetry, flavor } & & {\tt horizontal symmetry, flavor } \\
735:  {\bf symmetry, family } & & \\ 
736:  {\bf symmetry breaking  } & & {\tt symmetry breaking  } \\
737:  {\bf neutrino, mixing angle } & & {\tt neutrino, mixing angle } \\ 
738:  {\bf neutrino, right-handed } & & \\
739:  {\bf neutrino, Majorana } & & \\ 
740:  {\bf neutrino, flavor } & & \\
741:  {\bf flavor, 3  } & & \\
742:  {\bf neutrino, solar  } & & {\tt neutrino, solar  } \\
743:  {\bf neutrino, cosmic radiation  } & & \\
744:  {\bf neutrino, oscillation  } & & {\tt neutrino, oscillation  } \\
745:  {\bf numerical calculations  } & & {\tt numerical calculations  } \\
746: %
747: \end{tabular}
748: 
749: \newpage
750: \noindent
751: \subsection{\, hep-ph/0103176}
752: \label{app:hep-0103176}
753: % Appendix~\ref{app:hep-0103176}
754: 
755: \begin{center}
756: {\small\bf Title:} 
757: {\bf Neutrino kinetics in a magnetized dense plasma} \\ 
758: \end{center}
759: 
760: \vspace{-5mm}
761: 
762: \begin{abstract} 
763: The relativistic kinetic equations (RKE) for lepton 
764: plasma in the presence of a strong external magnetic field are
765: derived in Vlasov approximation. The new RKE for the electron spin
766: distribution function includes the weak interaction with neutrinos
767: originated by the axial vector current ($\sim c_A$) and provided by the
768: parity nonconservation. In a polarized electron gas Bloch equation
769: describing the evolution of the magnetization density perturbation is
770: derived from the electron spin RKE being modified in the presence of
771: neutrino fluxes. Such modified hydrodynamical equation allows to obtain
772: the new dispersion equation in a magnetized plasma from which {\it the
773: neutrino driven instability of spin waves} can be found. It is shown that
774: this instability is more efficient e.g. in a magnetized supernova than
775: the analogous one for Langmuir waves enhanced in an isotropic plasma. \\
776: \end{abstract} 
777: 
778: \begin{tabular}{lll}
779: \noindent
780: \underline{{\rm AUTEX Report:}} &  & \underline{{\rm SPIRES Report:}} \\ \\
781: {\bf kinetics, relativistic} & &  {\tt neutrino, kinematics} \\
782: {\bf lepton, plasma} & & {\tt lepton, plasma}  \\
783: {\bf plasma, magnetic} & &  \\ 
784: {\bf magnetic field, external field} & & 
785:                                  {\tt magnetic field, external field} \\
786: {\bf magnetic field, high} & & {\tt magnetic field, high} \\
787: {\bf Vlasov equation} & &  {\tt Vlasov equation} \\
788: {\bf electron, spin } & & {\tt electron, spin } \\
789: {\bf neutrino, interaction} & & {\tt neutrino electron, interaction} \\
790: {\bf current, axial-vector} & &  \\    
791: {\bf parity, violation} & &  \\   
792: {\bf electron, gas} & &  {\tt electron, gas} \\
793: {\bf electron, polarization} & & {\tt electron, polarization} \\
794: {\bf density, perturbation} & & {\tt density, perturbation}  \\
795: {\bf neutrino, beam} & & {\tt neutrino, beam} \\
796: {\bf neutrino, flux} & &  \\
797: {\bf dispersion relations} & &  {\tt dispersion} \\
798: {\bf stability} & &  {\tt stability} \\
799: %
800: 
801: {\bf astrophysics, supernova} & &  \\ 
802: {\bf lepton, current}  & & \\ 
803: {\bf current, conservation law}  & & \\ \hline
804:  & & {\tt neutrino, form factor} \\
805:  & & {\tt neutrino, charge} \\
806:  & & {\tt charge, induced} \\ \\
807: %
808: \end{tabular}
809: 
810: \newpage
811: \noindent
812: \subsection{\, hep-ph/0106085}
813: \label{app:hep-0106085}
814: 
815: \begin{center}
816: {\small\bf Title:} 
817: {\bf Models of neutrino masses and mixings} \\
818: \end{center}
819: 
820: \begin{abstract}
821: We briefly review models of neutrino masses and mixings. In view of 
822: the existing experimental ambiguities many possibilities are still
823: open. After an overview of the main alternative options we focus on 
824: the most constrained class of models based on three widely
825: split light neutrinos within SUSY Grand Unification. \\
826: \end{abstract}
827: 
828: \begin{tabular}{lll}
829: \noindent
830: \underline{{\rm AUTEX Report:}} &  & \underline{{\rm SPIRES Report:}} \\ \\
831: %
832: {\bf neutrino, mass} & & {\tt neutrino, mass} \\
833: {\bf neutrino, mixing angle} & &  \\
834: {\bf neutrino, flavor} & & {\tt neutrino, flavor}  \\
835: {\bf flavor, 3} & &  \\
836: {\bf flavor, 4} & &  \\ 
837: {\bf  mass, texture} & &  \\ 
838: {\bf hierarchy} & & {\tt hierarchy}  \\
839: {\bf lepton number, violation} & & {\tt flavor, violation}  \\
840: {\bf supersymmetry} & &  {\tt supersymmetry} \\
841: {\bf grand unified theory, SU(5)} & & {\tt grand unified theory, SU(5)} \\
842: {\bf grand unified theory, SO(10)} & & 
843:   {\tt grand unified theory, SO(10)}  \\ 
844: {\bf space-time, higher-dimensional} & &  
845:   {\tt space-time, higher-dimensional} \\
846: {\bf horizontal symmetry, U(1)} & & {\tt horizontal symmetry}  \\
847: {\bf charge, abelian} & &  \\
848: {\bf seesaw model} & &  \\   
849: {\bf neutrino, oscillation} & &  {\tt neutrino, oscillation} \\
850: {\bf neutrino, sterile} & &  {\tt neutrino, sterile} \\ 
851: {\bf neutrino, solar} & &  \\   
852: {\bf neutrino, right-handed } & &  \\  
853: {\bf neutrino, Dirac} & &  \\  
854: {\bf neutrino, Majorana } & &  \\  
855: {\bf MSW effect} & &  \\   
856: {\bf astrophysics, missing-mass } & &  \\  \\ \hline
857:       & &  {\tt neutrino, mass difference} \\
858:  %
859: \end{tabular}
860: 
861: \newpage
862: \noindent
863: \subsection{\, hep-ph/0107217}
864: \label{app:hep-0107217}
865: 
866: \begin{center}
867: {\small\bf Title:} 
868: {\bf Photon damping caused by electron-positron pair production in a 
869:      strong magnetic field} \\
870: \end{center}
871: 
872: \begin{abstract}
873:  Damping of an electromagnetic wave in a strong magnetic field 
874: is analyzed in the kinematic region near the
875: threshold of electron-positron pair production. Damping of the 
876: electromagnetic field is shown to be noticeably
877: nonexponential in this region. The resulting 
878: width of the photon $\gamma \to e^+ e^-$ decay is 
879: considerably smaller than previously known results. \\
880: \end{abstract}
881: 
882: \begin{tabular}{lll}
883: \noindent
884: \underline{{\rm AUTEX Report:}} &  & \underline{{\rm SPIRES Report:}} \\ \\
885: %
886:  {\bf photon, absorption} & & \\
887:  {\bf photon, leptonic decay} & & \\ 
888:  {\bf electron, pair production } & & 
889:    {\tt electron \underline{positron}, pair production}  \\ 
890: %   & & {\tt pair production} \\
891:  {\bf magnetic field, external field} & & 
892:    {\tt magnetic field, external field}\\
893:  {\bf magnetic field, high } & & {\tt magnetic field, high } \\
894:  {\bf photon, dispersion relations} & & \\ 
895:  {\bf field equations, solution} & & \\
896:  {\bf threshold, pair production} &  & {\tt threshold} \\ 
897:  {\bf kinematics } & & \\
898:  {\bf photon, polarization } & & \\
899:  {\bf vacuum polarization, ($\to$) } & & \\
900:  {\bf magnetic field } & & \\
901:  {\bf photon, width } & & \\
902:  {\bf photon $\to$ positron electron } & & \\ \\ \hline
903:        & & {\tt electromagnetic field} \\ 
904:        & & {\tt (0) photon, energy loss} \\ 
905:        & & {\tt numerical calculations} \\ 
906: %
907: \end{tabular}
908: 
909: \newpage
910: \noindent
911: \subsection{\, hep-ph/0112171}
912: \label{app:hep-0112171}
913: 
914: \begin{center}
915: {\small\bf Title:} 
916: {\bf Neutrino superbeam and factory tests of grand unified model
917: predictions for the large mixing angle and LOW solar neutrino solutions} \\
918: \end{center}
919: 
920: \begin{abstract}
921: Within the framework of an SO(10) GUT model that can accommodate both the 
922: LMA and LOW solar neutrino mixing solutions by appropriate choice of the
923: right-handed Majorana matrix elements, we present explicit predictions for the 
924: neutrino oscillation parameters $\Delta m^2_{21}$, $\sin^2 2\theta_{12}$, 
925: $\sin^2 2\theta_{23}$, $\sin^2 2\theta_{13}$, and $\delta_{CP}$.  Given the 
926: observed near maximality of the atmospheric mixing, the model favors the 
927: LMA solution and predicts that $\delta_{CP}$ is small.  The suitability of 
928: Neutrino Superbeams and Neutrino Factories for precision tests of the two 
929: model versions is discussed. \\
930: \end{abstract}
931: 
932: \begin{tabular}{lll}
933: \noindent
934: \underline{{\rm AUTEX Report:}} &  & \underline{{\rm SPIRES Report:}} \\ \\
935: %
936:  {\bf neutrino, mixing angle } & & {\tt neutrino, mixing angle } \\ 
937:  {\bf grand unified theory, SO(10)} & & {\tt grand unified theory, SO(10)} \\ 
938:  {\bf neutrino, solar} & & {\tt neutrino, solar} \\  
939:  {\bf neutrino, cosmic radiation} & & \\ 
940:  {\bf neutrino, right-handed} & & \\ 
941:  {\bf mass, Majorana} & & \\ 
942:  {\bf neutrino, oscillation} & & \\ 
943:  {\bf neutrino, mass difference} & & {\tt neutrino, mass difference} \\ 
944:  {\bf MSW effect} & & \\ 
945:  {\bf violation, CP} & & \\ 
946:  {\bf numerical calculations,} & & {\tt numerical calculations,} \\  
947:  {\bf interpretation of experiments} & & \\ \\ \hline
948:        & & {\tt neutrino, beam} \\     
949:        & & {\tt neutrino, particle source} \\     
950: %
951: \end{tabular}
952: 
953: \newpage
954: \noindent
955: \subsection{hep-ph/9910476}
956: \label{app:hep-9910476}
957: 
958: \begin{center}
959: {\small\bf Title:} 
960: {\bf Neutrino Oscillations in Electromagnetic Fields} \\
961: \end{center}
962: 
963: \begin{abstract}
964: Oscillations of neutrinos $\nu_L \leftrightarrow \nu_R$ in presence of 
965: an arbitrary electromagnetic field are considered. We introduce the
966: Hamiltonian for the neutrino spin evolution equation that accounts for
967: possible effects of interaction of neutrino magnetic $\mu$ and
968: electric $\epsilon$ dipole moments with the transversal
969: (in respect to the neutrino momentum) and also the longitudinal
970: components of electromagnetic field.  Using this Hamiltonian we
971: predict the new types of resonances in the neutrino oscillations 
972: $\nu_L \leftrightarrow \nu_R$ in the presence of the field of an 
973: electromagnetic wave and in combination of an electromagnetic wave
974:  and constant magnetic field. The possible influence of the 
975: longitudinal magnetic field on neutrino oscillations is emphasized. \\
976: \end{abstract}
977: 
978: \begin{tabular}{lll}
979: \noindent
980: \underline{{\rm AUTEX Report:}} &  & \underline{{\rm SPIRES Report:}} \\ \\
981: {\bf neutrino, oscillation} & & {\tt neutrino, oscillation} \\ 
982: {\bf neutrino, left-handed} & & \\ 
983: {\bf neutrino, right-handed} & & \\ 
984: {\bf electromagnetic field, external field } & & 
985:                       {\tt electromagnetic field} \\ 
986: {\bf electromagnetic field, longitudinal} & & \\ 
987: {\bf electromagnetic field, transversal} & & \\ 
988: {\bf magnetic field, longitudinal} & & \\ 
989: {\bf neutrino, momentum} & & \\ 
990: {\bf neutrino, spin} & & {\tt neutrino, spin} \\ 
991: {\bf effective Hamiltonian} & & \\  
992: {\bf neutrino, magnetic moment} & &  {\tt neutrino, magnetic moment} \\ 
993: {\bf neutrino, electric moment} & & \\ 
994: {\bf moment, dipole} & & \\ 
995: {\bf resonance} & & \\ 
996: \end{tabular}
997: 
998: \newpage
999: \noindent
1000: \subsection{\, hep-ph/9812408}
1001: \label{app:hep-9812408}
1002: % Appendix~\ref{app:hep-9812408}
1003: 
1004: \begin{center}
1005: {\small\bf Title:} 
1006: {\bf Field-induced axion decay $a \to e^+ e^-$ via plasmon} \\
1007: \end{center}
1008: 
1009: \begin{abstract}
1010: The axion decay $a \to e^+ e^-$ via a plasmon is investigated 
1011: in an external magnetic field. The results we have 
1012: obtained demonstrate a strong catalyzing influence of the field
1013: as the axion lifetime in the magnetic field of order $10^{15}$~G
1014: and at temperature of order 10~MeV is reduced to $10^{4}$~sec. \\
1015: \end{abstract}
1016: 
1017: \begin{tabular}{lll}
1018: \noindent
1019: \underline{{\rm AUTEX Report:}} &  & \underline{{\rm SPIRES Report:}} \\ \\
1020: {\bf axion, leptonic decay } & & {\tt axion, leptonic decay} \\ 
1021: {\bf electron, pair production} & & \\  
1022: {\bf magnetic field, external field} & & 
1023:                 {\tt magnetic field, external field} \\  
1024: {\bf yield, enhancement} & & \\   
1025: {\bf plasmon, propagator} & & \\  
1026: {\bf axion, dispersion relations}  & & {\tt dispersion relations} \\   
1027: {\bf plasmon, dispersion relations}  & & \\ 
1028: {\bf plasmon, longitudinal}  & & \\  
1029: {\bf axion, lifetime} & & {\tt axion, lifetime} \\  
1030: {\bf temperature } & & \\ 
1031: {\bf numerical calculations} & & {\tt numerical calculations} \\  
1032: {\bf Feynman graph}  & & \\  
1033: {\bf axion $\to$ positron electron } & & \\ \\ \hline
1034: {\sl coupling, (axion photon)} & & \\   
1035:   & &  {\tt plasma}    
1036: 
1037: %
1038: \end{tabular}
1039: 
1040: \vspace{7mm}
1041: 
1042: \noindent Note, that the keychain {\sl ``coupling, (axion photon)''} 
1043: is not present in the {\rm AUTEX Report}, but it will be included 
1044: in the next stage of the system development. 
1045: 
1046: \newpage
1047: \noindent
1048: \subsection{hep-ph/9710219}
1049: \label{app:hep-9710219}
1050: % Appendix~\ref{app:hep-9710219}
1051: 
1052: \begin{center}
1053: {\small\bf Title:} 
1054: {\bf Neutrino transitions $\nu \to \nu \gamma$,
1055:        $\nu \to \nu e^+ e^-$ in a strong magnetic field 
1056:        as a possible origin of cosmological $\gamma$-burst} \\
1057: \end{center}
1058: 
1059: \begin{abstract}
1060: The high energy neutrino transitions with the photon and
1061:  electron-positron pair creation in a strong magnetic field  in the
1062:  framework of the Standard Model are investigated. The process
1063:  probabilities and the mean values of the neutrino energy and momentum
1064:  loss are presented. The asymmetry of outgoing neutrinos, as a
1065:  possible source of sufficient recoil ``kick'' velocity of a remnant
1066:  and the emission of $e^+ e^-$-pairs and $\gamma$-quanta in a ``polar
1067:  cap'' region of a remnant, as a possible origin of cosmological
1068:  $\gamma$-burst are discussed. \\ 
1069: \end{abstract}
1070: 
1071: \begin{tabular}{lll}
1072: \noindent
1073: \underline{{\rm AUTEX Report:}} &  & \underline{{\rm SPIRES Report:}} \\ \\
1074: {\bf neutrino, transition} & & \\
1075: {\bf neutrino, radiative decay} & & {\tt neutrino, radiative decay} \\ 
1076: {\bf neutrino, leptonic decay} & & {\tt neutrino, leptonic decay} \\    
1077: {\bf electron, pair production} & & {\tt lepton, pair production} \\
1078: {\bf flavor, conservation law} & & \\                
1079: {\bf neutrino, relativistic} & & \\
1080: {\bf magnetic field, high } & & {\tt magnetic field, external field} \\  
1081: {\bf electroweak interaction} & & {\tt electroweak interaction} \\
1082: {\bf neutrino, energy loss} & & {\tt neutrino, energy loss} \\
1083: {\bf neutrino, energy-momentum} & & \\
1084: {\bf neutrino, emission} & & \\
1085: {\bf neutrino, momentum} & & \\
1086: {\bf momentum, asymmetry} & & \\
1087: {\bf astrophysics, supernova} & & {\tt astrophysics, supernova} \\
1088: {\bf n, matter} & & {\tt n, matter} \\
1089: {\bf parity, violation} & & \\ 
1090: {\bf pulsar, velocity} & & \\ 
1091: {\bf gamma ray burst} & & \underline{{\tt photon, cosmic radiation}} \\
1092: {\bf numerical calculations} & & \\   
1093: {\bf neutrino $\to$ neutrino photon} & & 
1094:                {\tt neutrino $\to$ neutrino photon} \\
1095: {\bf neutrino $\to$ positron electron } & & 
1096:                {\tt neutrino $\to$ positron electron } \\
1097: \hspace*{24mm} {\bf neutrino} & & \hspace*{26mm}{\tt neutrino} \\ \\\hline
1098:  & & {\tt (0) neutrino, massive} \\ 
1099: \end{tabular}
1100: 
1101: \newpage
1102: \noindent
1103: \subsection{\, hep-ph/9404289}
1104: \label{app:hep-9404289}
1105: % Appendix~\ref{app:hep-9404289}
1106: 
1107: \begin{center}
1108: {\small\bf Title:} 
1109: {\bf The Radiative decay of a high-energy neutrino in the Coulomb
1110:      field of a nucleus} \\
1111: \end{center}
1112: 
1113: \begin{abstract}
1114: In the framework of the Standard Model with lepton mixing the radiative
1115: decay $\nu_i \rightarrow \nu_j \gamma$ of a neutrino of high ($E_\nu \sim
1116: 100 \, GeV$) and super-high ($E_\nu \ge 1 \, TeV$) energy is investigated
1117: in the Coulomb field of a nucleus. Estimates of the decay probability and
1118: ``decay cross-section'' for neutrinos of these energies in the electric
1119: field of a nucleus permit one to discuss the general possibility of carrying
1120: out a neutrino experiment, subject to the condition of availability in the
1121: future of a beam of neutrinos of these high energies. Such an experiment
1122: could give unique information on mixing angles in the lepton sector of
1123: the Standard Model which would be independent of the specific neutrino
1124: masses if only the threshold factor ($1 - m_j^4 / m_i^4$) was not close
1125: to zero.
1126: \end{abstract}
1127: 
1128: \begin{tabular}{lll}
1129: \noindent
1130: \underline{{\rm AUTEX Report:}} &  & \underline{{\rm SPIRES Report:}} \\ \\
1131: {\bf neutrino, radiative decay} & & {\tt neutrino, radiative decay}\\  
1132: {\bf flavor, violation} & & \\  
1133: {\bf neutrino, relativistic} & & \\  
1134: {\bf neutrino, massive} & & \\  
1135: {\bf external field, Coulomb} & & \\  
1136: {\bf nucleus, electric field} & & {\tt nucleus, electric field} \\  
1137: {\bf electroweak interaction} & & {\tt electroweak interaction} \\  
1138: {\bf lepton, interference} & & {\tt lepton, interference} \\  
1139: {\bf lepton, mixing angle} & & {\tt lepton, mixing angle} \\  
1140: {\bf mass, threshold} & & \\  
1141: {\bf neutrino, beam} & & \\  
1142: {\bf cross section, decay} & & {\tt cross section, decay} \\  
1143: {\bf neutrino  $\to$ neutrino photon} & & \\  
1144: %
1145: \end{tabular}
1146: 
1147: \newpage
1148: \noindent
1149: \subsection{\, astro-ph/9812366}
1150: \label{app:astro-9812366}
1151: 
1152: \begin{center}
1153: {\small\bf Title:} 
1154: {\bf Neutrino emission due to Cooper pairing of nucleons in 
1155: cooling neutron stars} \\
1156: \end{center}
1157: 
1158: \begin{abstract}
1159: The neutrino energy emission rate due to formation
1160: of Cooper pairs of neutrons and protons in the superfluid
1161: cores of neutron stars is studied. The cases of singlet-state pairing 
1162: with isotropic superfluid gap and triplet-state
1163: pairing with anisotropic gap are analysed.
1164: The neutrino emission due to the singlet-state pairing of protons
1165: is found to be greatly suppressed with respect to the
1166: cases of singlet- or triplet-state pairings of neutrons.
1167: The neutrino emission due to pairing of neutrons is shown to 
1168: be very important in the superfluid neutron--star cores with the 
1169: standard neutrino luminosity and with the luminosity
1170: enhanced by the direct Urca process. It can greatly accelerate both, 
1171: standard and enhanced, cooling of neutron stars with superfluid cores.
1172: This enables one to interpret the data on surface temperatures of
1173: six neutron stars, obtained by fitting the observed spectra with the
1174: hydrogen atmosphere models, by the standard cooling with moderate 
1175: nucleon superfluidity.
1176: \end{abstract}
1177: 
1178: \begin{tabular}{lll}
1179: \noindent
1180: \underline{{\rm AUTEX Report:}} &  & \underline{{\rm SPIRES Report:}} \\ \\
1181: %
1182: {\bf neutrino, energy} & & \\
1183: {\bf neutrino, emission} & & {\tt neutrino, emission} \\
1184: {\bf astrophysics} & & {\tt astrophysics, model} \\
1185: {\bf n, matter} & &  {\tt n, matter} \\
1186: {\bf n, pair} & & \\
1187: {\bf p, pair} & & \\
1188: {\bf pair, Cooper} & & \\
1189: {\bf nucleon, superfluid} & & {\tt superfluid} \\
1190: {\bf superfluid, gap} & & \\
1191: {\bf anisotropy} & & \\
1192: {\bf neutrino, luminosity} & & {\tt neutrino, luminosity} \\
1193: {\bf Urca process} & & \\
1194: {\bf star, energy loss} & & \\
1195: {\bf temperature, surface} & & \\
1196: {\bf critical phenomena, temperature} & & \\
1197: {\bf redshift} & & \\
1198: {\bf temperature, dependence} & & {\tt temperature, dependence} \\
1199: {\bf bremsstrahlung} & & \\
1200: {\bf numerical calculations} & & 
1201:   {\tt numerical calculations,} \\
1202:  & &  {\tt interpretation of experiments} \\
1203: %
1204: \end{tabular}
1205: 
1206: 
1207: \end{document}
1208: