チケット #220: help_en.2.html

ファイル help_en.2.html, 21.4 KB (登録者: yy, 16 年 前)

英語版ヘルプページ rev 2

行番号 
1<!DOCTYPE html
2        PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
3         "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
4<html xmlns="http://www.w3.org/1999/xhtml" lang="en" xml:lang="en">
5<head>
6<title>OReFiL: Help</title>
7<style type="text/css">
8<!--/* <![CDATA[ */
9.a {color: green}
10.b {border: thin dotted #888; font-size:18px}
11.e {background: #eee}
12.m {text-decoration: none; background: #eef; margin: .1ex .1em .1ex .1em}
13.n {background: #ddd}
14a.m:link {color: #00f}
15a.m:visited {color: #00f}
16a.m:hover {color: #770}
17a.m:active {color: #00f}
18h1 {text-indent: .2em; border-color: black; border-width: 0 0 1px 4px; border-style: none none solid double}
19h2 {text-indent: .2em; border-color: black; margin: 0}
20h3 {padding: .2em .3em .2em .3em; margin: 0 0 .5em 0; color: white; background: #444}
21ul {margin-top: 0; margin-bottom: 3ex}
22li {margin-bottom: .7ex}
23body {font-family: Arial, sans-serif; margin: 1.5ex; padding: 0; line-height: 1.4; color: #333; font-size: 0.9em}
24html {overflow-y:scroll;}
25p {margin: .7ex .7em .7ex .7em}
26/* ]]> */-->
27</style>
28</head>
29<body>
30<h1>OReFiL Help</h1>
31<ul>
32<li><a href="#sec_1">What is OReFiL?</a></li>
33<li><a href="#sec_2">What is the benefit of using OReFiL?</a></li>
34<li><a href="#sec_3">Why is the title ... ?</a></li>
35<li><a href="#sec_4">How to compose a query?</a></li>
36<li><a href="#sec_5">How to use modifiers?</a></li>
37<li><a href="#sec_6">How to use operators?</a></li>
38<li><a href="#sec_7">How to see the result &amp; what can I do next?</a></li>
39<li><a href="#sec_8">What is `Permalink`s shown up on a search result?</a></li>
40<li><a href="#sec_9">What is the relevance feedback?</a></li>
41<li><a href="#sec_10">Why does not my work show up?</a></li>
42<li><a href="#sec_11">How to change the order of results?</a></li>
43<li><a href="#sec_12">What sort options can I use?</a></li>
44<li><a href="#sec_13">What is the "Show NEW entries only"?</a></li>
45<li><a href="#sec_14">What is the benefit of signing in?</a></li>
46<li><a href="#sec_15">How to obtain an account to sign in?</a></li>
47<li><a href="#sec_16">How to sign in?</a></li>
48<li><a href="#sec_17">How to sign out?</a></li>
49<li><a href="#sec_18">How to rate an online resource?</a></li>
50<li><a href="#sec_19">How to cancel a rating I did?</a></li>
51<li><a href="#sec_20">How to make a comment on an online resource?</a></li>
52<li><a href="#sec_21">How to modify my comment?</a></li>
53<li><a href="#sec_22">How to delete my comment?</a></li>
54<li><a href="#sec_23">How to read comments submitted by others?</a></li>
55<li><a href="#sec_24">How to narrow down MeSH terms shown in the box at the right hand side?</a></li>
56<li><a href="#sec_25">How to subscribe a feed of a search result?</a></li>
57<li><a href="#sec_26">How to obtain a list of newly indexed online resources?</a></li>
58<li><a href="#sec_27">How to obtain PubMed entries for an online resource?</a></li>
59<li><a href="#sec_28">How to get newly added PubMed entries for an online resource?</a></li>
60<li><a href="#sec_29">Can't I obtain an update time of each entry?</a></li>
61<li><a href="#sec_30">How to obtain a search result through WebAPI?</a></li>
62<li><a href="#sec_31">How to obtain an online resource information through WebAPI?</a></li>
63<li><a href="#sec_32">How to change the language?</a></li>
64<li><a href="#sec_33">I cannot use the OReFiL SOAP API from Taverna.</a></li>
65<li><a href="#sec_34">Who does develop/maintain this system?</a></li>
66</ul>
67<center><p><a href="index.cgi">Begin a search</a></p>
68<p>Comments and suggestions are welcome, please send an email to yayamamo AT cb DOT k DOT u-tokyo DOT ac DOT jp.</p></center>
69<h3><a id="sec_1">What is OReFiL?</a></h3>
70<p>OReFiL stands for <i><b>O</b>nline <b>Re</b>source <b>Fi</b>nder for <b>L</b>ifescience</i>.
71It facilitates a search for online resources that are introduced in peer-reviewed papers.
72You can search by MeSH terms or author names in addition to free words.
73OReFiL extracts all URLs from MEDLINE abstracts and PubMed-indexed BioMed Central full-papers (implementation or availability sections), and indexes them with MeSH terms and author names.</p>
74<h3><a id="sec_2">What is the benefit of using OReFiL?</a></h3>
75<ol>
76<li>OReFiL returns up-to-date query-relevant online resources that is introduced in peer-reviewed papers.</li>
77<li>Users can search for online resources not only by free words, but also by MeSH terms or author names.</li>
78<li>Users can easily verify each hit resource by following the links to its corresponding PubMed entry, webpages having a link to that of the hit resource, or papers citing it through the search systems of BioMed Central, Scirus, HighWire Press, or Google Scholar.</li>
79<li>Users can quickly confirm the existence of an online resource webpage.</li>
80</ol>
81<h3><a id="sec_3">Why is the title ... ?</a></h3>
82<ul>
83<li><b>Page was not found</b></li>
84<li style="list-style: none; display: inline">
85<p>This means our crawler got the HTTP status code of 404 when accessing to the webpage.
86The probability of its removal is high, although it might be temporal.
87You can hide those entries by checking the "Hide page-not-found entries" box just below the query box.</p>
88</li>
89<li><b>Could not get the data</b></li>
90<li style="list-style: none; display: inline">
91<p>This means our crawler could not fetch the webpage when accessing to the URL (the HTTP status code was 5xx).
92Several cases can be the reason; the server was too busy or down, the net was congested, or the server malfunctioned.
93So, it might be accessible when you click the link.</p>
94</li>
95<li><b>Could not obtain the title</b></li>
96<li style="list-style: none; display: inline">
97<p>This means our crawler could not find a webpage's title or that of what looks like.
98At least the webpage existed when our crawler accessed.</p>
99</li>
100<li><b>Could not fetch the page: xxx ...</b></li>
101<li style="list-style: none; display: inline">
102<p>This means our crawler could not fetch the webpage when accessing to the URL for a reason not listed above.
103The HTTP status code and the corresponding message follows.</p>
104</li>
105</ul>
106<h3><a id="sec_4">How to compose a query?</a></h3>
107<p>A query can be any words consisting of alphanumerical characters ("a" - "z", "A" - "Z", and "0" - "9"), "-", or "_".
108So, any punctuation is ignored.
109OReFiL supports not only a boolean-based search, but also a language-model based one, which utilizes probabilistic inference.
110Boolean-based and language-model based searches are mutually exclusive.
111If you type in words without a boolean operator such as <i>#band</i> (described <a href="#sec_6">below</a>), the search is language-model based, and vice versa.
112For example, a search to the following query is a language-model based:
113<i>apoptosis caspase</i>, and the following one is a boolean based:
114<i><a href="http://orefil.dbcls.jp/search?query=%23band%28apoptosis+caspase%29&amp;count=50&amp;sort_by=relevance&amp;hide_unfetched=on">#band(apoptosis caspase)</a></i>.
115Any word can be modified using modifiers such as <i>.mesh</i> (described <a href="#sec_5">below</a>) to specify a context where that word appears such as in an author name or a title/abstract.</p>
116<h3><a id="sec_5">How to use modifiers?</a></h3>
117<p>The following modifiers can be used to specify a context.
118Each modifier can be added just after a word with the delimiter '.' (period).
119For example, if you want OReFiL to search for any entries whose MEDLINE abstracts (including titles) contain the word <i>pathway</i>, you can add the modifier <i>tiab</i>, so, the query is <b><i><a href="http://orefil.dbcls.jp/search?query=pathway.tiab&amp;count=50&amp;sort_by=relevance&amp;hide_unfetched=on">pathway.tiab</a></i></b>.</p>
120<ul>
121<li><i>w</i>.tiab</li>
122<li style="list-style: none; display: inline">
123<p>Word <i>w</i> is in MEDLINE titles and abstracts.
124(Ex. <i><a href="http://orefil.dbcls.jp/search?query=alzheimer.tiab&amp;count=50&amp;sort_by=relevance&amp;hide_unfetched=on">alzheimer.tiab</a></i>)</p>
125</li>
126<li><i>w</i>.auth</li>
127<li style="list-style: none; display: inline">
128<p>Word <i>w</i> is in MEDLINE author lists.
129(Ex. <i><a href="http://orefil.dbcls.jp/search?query=john.auth&amp;count=50&amp;sort_by=relevance&amp;hide_unfetched=on">john.auth</a></i>)</p>
130</li>
131<li><i>w</i>.mesh</li>
132<li style="list-style: none; display: inline">
133<p>Word <i>w</i> is in annotated MeSH terms regardless of whether they are MEDLINE entry's descriptors of major topics or not.
134(Ex. <i><a href="http://orefil.dbcls.jp/search?query=p53.mesh&amp;count=50&amp;sort_by=relevance&amp;hide_unfetched=on">p53.mesh</a></i>)</p>
135</li>
136<li><i>w</i>.majr</li>
137<li style="list-style: none; display: inline">
138<p>Word <i>w</i> is in annotated MeSH terms that are MEDLINE entry's descriptors of major topics or their conceptual ancestors in the MeSH hierarchy.
139(Ex. <i><a href="http://orefil.dbcls.jp/search?query=p53.majr&amp;count=50&amp;sort_by=relevance&amp;hide_unfetched=on">p53.majr</a></i>)</p>
140</li>
141<li><i>w</i>.noexp</li>
142<li style="list-style: none; display: inline">
143<p>Word <i>w</i> is in annotated MeSH terms that are MEDLINE entry's descriptors of major topics.
144(Ex. <i><a href="http://orefil.dbcls.jp/search?query=p53.noexp&amp;count=50&amp;sort_by=relevance&amp;hide_unfetched=on">p53.noexp</a></i>)</p>
145</li>
146</ul>
147<h3><a id="sec_6">How to use operators?</a></h3>
148<p>OReFiL uses Lemur Toolkit (Indri), and the query language can be learned <a href="http://www.lemurproject.org/lemur/IndriQueryLanguage.php">here</a> comprehensively.
149Here, I just pick up only three of the rich operators that might be frequently used.
150A basic rule is that many operators begin with "#" and each operator takes arguments within brackets.
151Combination can be accepted such as <i><a href="http://orefil.dbcls.jp/search?query=%23band%28metabolic+pathway+%231%28Computational+Biology%29.majr%29&amp;count=50&amp;sort_by=relevance&amp;hide_unfetched=on">#band(metabolic pathway #1(Computational Biology).majr)</a></i>.</p>
152<ul>
153<li>#combine</li>
154<li style="list-style: none; display: inline">
155<p>This is the default operator when you just type in words in the query box without any operators.
156OReFiL searches for query-relevant entries in terms of the language model.
157Multiple words can be accepted and the word order may matter.
158An example is <i><a href="http://orefil.dbcls.jp/search?query=%23combine%28protein+protein+interaction%29&amp;count=50&amp;sort_by=relevance&amp;hide_unfetched=on">#combine(protein protein interaction)</a></i>.</p>
159</li>
160<li>#band</li>
161<li style="list-style: none; display: inline">
162<p>This is the boolean "AND" operator.
163An example is <i><a href="http://orefil.dbcls.jp/search?query=%23band%28apoptosis+caspase%29&amp;count=50&amp;sort_by=relevance&amp;hide_unfetched=on">#band(apoptosis caspase)</a></i>.</p>
164</li>
165<li>#filreq</li>
166<li style="list-style: none; display: inline">
167<p>This operator takes two arguments and means "filter require",
168that is, OReFiL searches for entries that match the first argument and ranks them according to second argument as a query.
169An example is <i><a href="http://orefil.dbcls.jp/search?query=%23filreq%28Software.noexp+%231%28sequence+analysis%29%29&amp;count=50&amp;sort_by=relevance&amp;hide_unfetched=on">#filreq(Software.noexp #1(sequence analysis))</a></i>.
170In this example, OReFiL only considers those resources whose MEDLINE entries have the MeSH term "Software" and rank them according to the query <a href="http://orefil.dbcls.jp/search?query=%231%28sequence+analysis%29&amp;count=50&amp;sort_by=relevance&amp;hide_unfetched=on">#1(sequence analysis)</a>.
171</p>
172</li>
173<li>#1</li>
174<li style="list-style: none; display: inline">
175<p>This is to be used for OReFiL to regard multiple words as a collocation.
176You can use this operator regardless of the search models (<i>i.e.</i>, boolean based or language-model based).
177For example, if you want to search for entries that contain the expression "signaling pathway", you need to compose the query <i><a href="http://orefil.dbcls.jp/search?query=%231%28signaling+pathway%29&amp;count=50&amp;sort_by=relevance&amp;hide_unfetched=on">#1(signaling pathway)</a></i>.
178</p>
179</li>
180</ul>
181<h3><a id="sec_7">How to see the result &amp; what can I do next?</a></h3>
182<ul>
183<li><b>MeSH terms</b> (encircled by a dotted line)</li>
184<li style="list-style: none; display: inline">
185<p>
186Those annotated to the MEDLINE abstracts included in the hit list are displayed in the alphabetical order.
187Each font size of them reflects its frequency in the hit list, so, those terms more frequently annotated in the MEDLINE abstracts in the hit list are bigger, and vice versa.
188They are only descriptors of major topics.
189Those MeSH terms also function as a filter to narrow down the search result.
190By clicking one of them, the browser automatically composes and fills in a required query appropriately.
191Users can easily add or remove a MeSH term in the query box by clicking the term in the MeSH term box;
192you do not need to type in a complicated query to narrow down.
193After confirming the query box, just clicking the "SUBMIT" button is enough to get the result.
194</p>
195</li>
196<li><b>MeSH</b></li>
197<li style="list-style: none; display: inline">
198<p>For each hit entry, MeSH terms appear below the title of the webpage and its URL.
199They are those annotated to the corresponding MEDLINE abstract(s).
200They are descriptors of the major topics.</p>
201</li>
202<li><i><b>Pages</b> linking to this site found by ...</i></li>
203<li style="list-style: none; display: inline">
204<p>These links are for checking a kind of "page rank" of the hit webpage, that is, by clicking a link, you can see how many webpages link to the hit webpage.
205Google and AltaVista provide the <i>link:</i> operator, by which you can get the list of the webpages that have a link to a webpage specified just after the operator.
206For example, you can search for webpages that have a link to PubMed by filling in the query <i>link:www.pubmed.gov</i>.
207The assumption is that those webpages to which more webpages have a link are more reputable.</p>
208</li>
209<li><i><b>Papers</b> citing this site found by ...</i></li>
210<li style="list-style: none; display: inline">
211<p>These links are similar to page rank but here each search service (<i>i.e.</i>, BioMed Central, Scirus, HighWire Press, and Google Scholar) returns a list of the papers that refer to the hit webpage.
212So, you can quickly search for those papers, for example, using the tool introduced in the hit webpage.</p>
213</li>
214<li><b>PubMed</b> entries</li>
215<li style="list-style: none; display: inline">
216<p>The title of the papers in which (title/abstract or BMC full-papers) the hit webpage's URL appears.
217Each is a link to the corresponding PubMed entry.</p>
218</li>
219</ul>
220<h3><a id="sec_8">What is `Permalink`s shown up on a search result?</a></h3>
221<p>This is a link to the detailed information of an entry.<br />
222In addition to the information on a hit list page, you can read comments posted by other users.</p>
223<h3><a id="sec_9">What is the relevance feedback?</a></h3>
224<p>To be precise, I use blind or "pseudo" relevance feedback here.
225It is a method to raise the ranks of relevant documents otherwise could not be raised, whose usefulness has been empirically proven. When searching for relevant documents, the system carries out a search process twice, that is, after conducting a first normal search, it searches again utilizing the first result.</p>
226<h3><a id="sec_10">Why does not my work show up?</a></h3>
227<p>It seems because you did not write down a URL in a title/abstract or a PubMed-indexed BMC paper.
228Even though you wrote down there and the MEDLINE entry does not show up here, we'd be glad if you'd inform us.
229Also, we're happy if you'd tell us pairs of a URL and a PubMed ID along with the pdf file for which you feel good to be indexed here and that have not been yet.
230The email address is yayamamo AT cb DOT k DOT u-tokyo DOT ac DOT jp, thanks.</p>
231<h3><a id="sec_11">How to change the sort order?</a></h3>
232<p>There are two ways as follows.</p>
233<ol><li>
234Just below the query box there are radio boxes showing `Relevance`, `Updated`, and `Rating`.
235After choosing one of them, just click the `SUBMIT` button.</li>
236<li>On the right hand of the header showing `Search Result for &#34;...&#34;`, there are two links of the three options `Relevance`, `Updated`, and `Rating`.
237By clicking either one, you can change the order.</li></ol>
238<h3><a id="sec_12">What sort options can I use?</a></h3>
239<p>There are three options as follows.</p>
240<ul><li>Relevance</li>
241<li>Updated</li>
242<li>Rating</li></ul>
243<h3><a id="sec_13">What is the "Show NEW entries only"?</a></h3>
244<p>Checking this box means that only newly added resources are on a result list.</p>
245<h3><a id="sec_14">What is the benefit of signing in?</a></h3>
246<p>You can rate or make a comment on a resource if you sign in.</p>
247<h3><a id="sec_15">How to obtain an account to sign in?</a></h3>
248<p>You need a DBCLS OpenID to sign in OReFiL, which can be obtained at <a href="https://openid.dbcls.jp/account/signup">https://openid.dbcls.jp/account/signup</a>.</p>
249<h3><a id="sec_16">How to sign in?</a></h3>
250<p>You need to input your DBCLS OpenID in the box at the right hand side of the `DBCLS OpenID URL:` on the top gray area.
251After that, just click the `Sign In` button.
252You may be required to input your password by the DBCLS OpenID server.</p>
253<h3><a id="sec_17">How to sign out?</a></h3>
254<p>Just click the `Sign Out` button on the top gray area.</p>
255<h3><a id="sec_18">How to rate an online resource?</a></h3>
256<p>First you need to sign in.
257After that, you can see an item `Your Rate:` in each hit entry or each resource's detailed information page.
258By clicking on a star you want, your can rate the resource.</p>
259<h3><a id="sec_19">How to cancel a rating I did?</a></h3>
260<p>Clicking the `x` button deletes your rate.</p>
261<h3><a id="sec_20">How to comment on an online resource?</a></h3>
262<p>First you need to sign in.
263After that, you can see an anchor `Add comment ...` just after the stars in each hit entry or each resource's detailed information page.
264Click the anchor and a comment box show up, where you can make a comment on the resource.
265Submission is done by clicking the `SUBMIT` button.</p>
266<h3><a id="sec_21">How to modify my comment?</a></h3>
267<p>Click the anchor `Edit comment ...` just below the comments on a resource.</p>
268<h3><a id="sec_22">How to delete my comment?</a></h3>
269<p>Click the anchor `Edit comment ...` just below the comments on a resource.
270Clear off all the words in the box, and click the `SUBMIT` button.</p>
271<h3><a id="sec_23">How to read comments submitted by others?</a></h3>
272<p>Open a resource's detailed information page.
273You can read those comments made by others at the bottom of the page.</p>
274<h3><a id="sec_24">How to narrow down MeSH terms shown in the box at the right hand side?</a></h3>
275<p>Input a word you want, and only MeSH terms that have a substring of it can be shown.</p>
276<h3><a id="sec_25">How to subscribe a feed of a search result?</a></h3>
277<p>You can obtain a search result by Atom or RSS (0.91).
278Widely used browsers such as Internet Explorer or FireFox can get those feeds.
279In addition, you can get by the following ways.</p>
280<ul><li>via Atom format: /search.atom?query=genome</li>
281<li>via RSS format: /search.rss?query=genome</li></ul>
282<h3><a id="sec_26">How to obtain a list of newly indexed online resources?</a></h3>
283<p>Newly added resources can be obtained via Atom/RSS feeds by the following ways.</p>
284<ul><li>via Atom format: /search/new.atom</li>
285<li>via RSS format: /search/new.rss</li></ul>
286<h3><a id="sec_27">How to obtain PubMed entries for an online resource?</a></h3>
287<p>PubMed entries for a resource can be obtained via Atom/RSS feeds by the following ways.</p>
288<ul><li>via Atom format: /online_resource/&lt;ResourceID&gt;.atom</li>
289<li>via RSS format: /online_resource/&lt;ResourceID&gt;.rss</li></ul>
290<h3><a id="sec_28">How to get newly added PubMed entries for an online resource?</a></h3>
291<p>Newly added PubMed entries for a resource can be obtained via Atom/RSS feeds by the following ways.</p>
292<ul><li>via Atom format: /online_resource/new/&lt;ResourceID&gt;.atom</li>
293<li>via RSS format: /online_resource/new/&lt;ResourceID&gt;.rss</li></ul>
294<h3><a id="sec_29">Can't I obtain an update time of each entry?</a></h3>
295<p>RSS version 0.91 cannot provide an update time for each entry.
296Use Atom that can provide if you need.</p>
297<h3><a id="sec_30">How to obtain a search result through WebAPI?</a></h3>
298<p>OReFiL provides REST/SOAP WebAPIs.
299Please see <a href="search_api_en.html">Specification of the Web APIs</a>.</p>
300<h3><a id="sec_31">How to obtain an online resource information through WebAPI?</a></h3>
301<p>OReFiL provides REST/SOAP WebAPIs.
302Please see <a href="online_resource_api_en.html">Specification of the Web APIs</a>.</p>
303<h3><a id="sec_32">How to change the language?</a></h3>
304<p>You can change the language by the following ways.</p>
305<ul><li>Click the anchor at the upper right of the top page.</li>
306<li>Insert a two letter language code (en:English or ja:Japanese) just after the host name (For example, `/en/search?query=genome`).</li></ul>
307<p>If there is no language code in the URL, OReFiL determines depending on you browser's preference (ACCEPT_LANGUAGE).</p>
308<h3><a id="sec_33">I cannot use the OReFiL SOAP API from Taverna.</a></h3>
309<p>Taverna 1.7.0 has a problem of handling SOAP messages and cannot utilize the OReFiL SOAP API.
310Please use a later version.</p>
311<h3><a id="sec_34">Who does develop/maintain this system?</a></h3>
312<p>OReFiL has been developed and maintained by Yasunori Yamamoto, PhD at <a href="http://dbcls.rois.ac.jp">Database Center for Life Sciences</a> (I was formerly at <a href="http://www.cb.k.u-tokyo.ac.jp/takagilab/en/index.html">Takagi Laboratory</a>, <a href="http://www.k.u-tokyo.ac.jp/index.html.en">Department of Computational Biology</a>, <a href="http://www.u-tokyo.ac.jp/index_e.html">University of Tokyo</a>).
313</p>
314</body>
315</html>