Viewing 6 reply threads
  • Author
    Posts
    • #52272
      Anastasios Asimakopoulos
      Keymaster
          @anastasios

          Here, you can share your findings. Of course, you can also add any reflections or questions too, as we are more than happy to help you understand the tools a little more. Remember you can also add a sceenshot of your results by using the ‘Insert image’ button when you are posting.

        • #53095
          Paula Acejo Cantero
          Participant
              @powla

              Hello,

              I used the economics corpora and I wanted to explore the use of “time” in this domain…. after all, time is money!

              First thing I did was to investigate the collocation of “time is…” so I used collocations and 1R as setting. I was surprised to find different as the first collocate was different and the second collocate is “not” which is also rather interesting. After that, the collocates are past participles used to describe patterns or point out to results (provided, given, shown…). I looked at what time is not since “not” was the 2nd top collocate. And found uniform, constant.

              Secondly, I investigated the top prepositions after time using “Cluster” (in, of, to, for) and using collocates for each “time in” “time of” etc, I found some patterns such as time to + verb, time of + noun, time + in+ -ing…

              Finally, I used the function “Cluster” for specialised vocabulary with 2 as cluster size to focus on bigrams. I ue the word cloud (attached). I adjusted the frequency and range to be a minimum of 10 each to focus on the most relevant pairs. From there, I selected the multiwords that had a compound structure of noun + noun being time the first noun, but acting syntactically as an adjective. These were: time series, time period, time step, time evolution, time scale, time window, time lag, time span, time interval & time horizon.

              Then, I wanted to extract examples in context for each and realised that KWIC sorted by frequency was given me rather technical texts compared to sorting it by value. Using the KWIC and sort by value, I can select sentences for each of these pairs for the students to examine and infer what the differences might be between them in use.  The target audience for this activity would be advanced learners of English for Specific Purpose or other students of that same level interested in this domain.

               

              Attachments:
              You must be logged in to view attached files.
            • #53097
              Paula Acejo Cantero
              Participant
                  @powla

                  Attaching the word cloud as it dis not workk before

                  • #53125
                    Anastasios Asimakopoulos
                    Keymaster
                        @anastasios

                        Thank you so much @powla for sharing the results of your mini-investigation of time in the Economics corpus. I particularly like your use of Clusters and the exploration of time + noun phrases as these can be very useful for international students. I usually ask my students to examine the top 5-10 bigrams and decide if they are bigrams indeed or whether the two words are part of other phrases.

                    • #53197
                      Siti Asmiyah
                      Participant
                          @siti-asmiyah

                          I am interested on the feature of Plot in Computational Linguistics. I first created the corpora of Computational Linguistics. Then I used the feature of Plot to search for the dispersion of the words ‘pattern’ so see the frequency and how important is the word in the corpora. To do this I uploaded the corpora into Antconc. Then, I went to plot and have the word ‘pattern’ in search query and pushed the start button. I then wanted to see how frequent the word ‘computation’ (only once across corpora) is used in the corpora. I selected the icon of overlay and have the word computation in blue color; while I previously had the word ‘pattern’ in red color. I found that although the corpora is about computational linguistics, the word pattern is used more frequently compared to the word ‘computation.’ Then I wanted to see when in the file the words are located. When I clicked on the colored stripes then I was generated to the source file. I was a bit curious why the colored words displayed on the file were only two.  My trial and errors and exploration then brought me to an understanding that if I clicked the hit icon and have more number on the hit, I would find more colored words in the file. The result of the plot search is displayed in the attached .png file.

                          Attachments:
                          You must be logged in to view attached files.
                        • #53225
                          Anastasios Asimakopoulos
                          Keymaster
                              @anastasios

                              Thank you for sharing your exploration of Plot @siti-asmiyah That’s correct, the lines in Plot represent the frequency of your search term in the corpus e.g. pattern* (patterns/patterns) occurs in 70 (1) out of 94 texts (2). This is what we call Range2, but it is a very crude dispersion measure. You might be interested in the distribution of the search term in terms of how many times it occurs in each file (3) as well as where in the file it occurs (4). I usually look at the column Dispersion (5), which by default is Juilland’s D. This number ranges from 0 to 1, with 0 indicating an uneven distribution and 1 an even distribution. So, the closer to 1 this value is, the more distributed the term is in this corpus file. I hope this helps a little understand the bars in the plot.

                               

                            • #53425
                              ZHEMIN CHEN
                              Participant
                                  @tina

                                  Thank you for your instructions and I benefit a lot from the course. Now I am trying to start my own research. I am interested in how academic writers give examples in different sections such as introduction and methods. The corpora created in Task 4 (AntcorGen Guide) are used to answer the question. What are the most frequent forms of exemplification in introduction and method sections?

                                  Could you please advise me on following:

                                  1.     How to search a list marker of exemplifications such as for example, such as and for instance? I tried many times, but it doesn’t work.

                                  2. I am still not sure how to compare the difference. Is it necessary to present the statistic of Likelihood and Effect?

                                  For example occurs 88 times in 47 out of 200 texts in the Introduction corpus, while 30 times in 18 out of 200 texts in the Method corpus.

                                  3.Would it be possible to use Keyword tool to compare the difference of exemplification markers? I also noticed there is another a Search Query list in Keyword tool and I don’t know how to use it.

                                  4. Could you share me how to study the concordance lines and identifiy the communicative patterns? It is difficult for me to identify the communicative patterns when I read the concordance lines and I don’t know how to start.

                                  I look forward to your respond and appreciate your time.

                                   

                                  Attachments:
                                  You must be logged in to view attached files.
                                • #60697
                                  Ian Barker
                                  Participant
                                      @ian

                                      I wanted to use the tools to see how the word ‘analysis’ is used in terms of KWIC and also as a cluster.  I don’t have any specific EAP discipline specific classes at the moment so I chose a very common core EAP word to examine actual patterns of use.  I choose to contrast Environmental Chemistry with Financial Management.  I used a sample size of 200 hits and focused upon ‘results and discussion’ section only.  1 mistake I made was to put a newly created database in the wrong folder – I was surprised to see ‘soil analysis’ showing up under Financial management – the importance of clear labelling here.

                                      For ease of contrast I decided to use KWIC (sorted right) and also 5 word clusters, for each of the 2 disciplines.  I wasn’t expecting hugely different patterns but I now feel that I can make a mini-corpus and use some lexical tools to utlise it. I can see the use of it in further EAP work, be it more multi-discipline classes or more discipline specific.  I did identify some common patterns to both disciplines and you would expect some rather different subject specific lexis to appear.

                                      For both disciplines the KWIC analysis showed that the adjective+KWIC+of+the +noun phrase was very common.  In Environmental chemistry the 3 most common nouns were  results/contribution of/distribution while Finance showed that effectiveness/number of /differences were common.   Rather elementary but I feel that with pre-sessional students new to Academic writing such simple patterns could be used in class to make them more of aware of how core EAP words are actually used in context – more likely to be of use  with more discipline specific classes.

                                       

                                      The 5 word cluster check showed that for both disciplines the most common phrases were Analysis was at the national / Analysis are presented in table.  After that, the results did rather differ – Chemistry – Analysis hif alpha gene notch / in soils and sediment  and Finance – Analysis due to incomplete cost / from the moh perspective.  This function would be very useful with a more discipline specific class.

                                      I’d like to use this with a pre-sessional class – identify some key lexis/structural patterns that they are struggling with and use this to create tasks.

                                       

                                       

                                       

                                      Attachments:
                                      You must be logged in to view attached files.
                                  Viewing 6 reply threads
                                  • You must be logged in to reply to this topic.